Intraoperative Imaging PP 169-174

FHB
(Radiology golden books)

Acta Neurochirurgica
Supplements
Editor: H.-J. Steiger

Intraoperative Imaging
Edited by
M. Necmettin Pamir, Volker Seifert, Talat Kırış
Acta Neurochirurgica
Supplement 109
SpringerWienNewYork
M. Necmettin Pamir
Professor and Chairman, Department of Neurosurgery, Acibadem University, School of Medicine,
Inonu Cad, Okur Sok 20, 34742 Kozytagi, Istanbul, Turkey
Volker Seifert
Univ. Klinikum Frankfurt, Zentrum Neurologie und Neurochirurgie, Klinik für Neurochirurgie,
Schleusenweg 2 16, 60528 Frankfurt, Haus 95, Germany
Talat Kırış
University of Istanbul, School of Medicine, Dept. Neurosurgery, 34390 Capa, Istanbul, Turkey
This work is subject to copyright.

All rights are reserved, whether the whole or part of the material is concerned, specifically those of translation, reprinting, re use
of illustrations, broadcasting, reproduction by photocopying machines or similar means, and storage in data banks.
Product Liability: The publisher can give no guarantee for all the information contained in this book. This does also
refer to information about drug dosage and application thereof. In every individual case the respective user must
check its accuracy by consulting other pharmaceutical literature. The use of registered names, trademarks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
# 2011 Springer-Verlag/Wien
Printed in Germany
SpringerWienNewYork is part of Springer Science+Business Media
springer.at
Typesetting: SPI, Pondichery, India
Printed on acid free and chlorine free bleached paper

SPIN: 12744831
Library of Congress Control Number: 2010935184
With 117 (partly coloured) Figures
ISSN 0065 1419

ISBN 978 3 211 99650 8 e ISBN 978 3 211 99651 5
DOI: 10.1007/978 3 211 99651 5
SpringerWienNewYork
Preface
In the pursuit of the goal of continuous improvement in surgical results, intraoperative

imaging technologies have taken an ever-increasing role in the daily practice of neurosur-
geons. To adapt available imaging technologies to the operating room a considerable amount
of effort has been focused on the subject. Most centers have taken individual and independent
approaches on the subject and an ever-diversifying field of “intraoperative imaging’’ has been
created. In an initiative of coordinating and symbiotically integrating these novel technolo-
gies, the international “Intraoperative Imaging Society’’ has been formed. After the second
international meeting of the society, this book is aimed to bring together both the essence and
details of the current status.
The initial drive for intraoperative imaging in neurosurgery came from the demands of
neurooncology. Accumulating evidence over the years has indicated that a more complete
resection of brain tumors was associated with a lower incidence of recurrences and longer
survival. This led to a search for techniques and technologies to improve the extent of surgical
resections. Stereotactic techniques have led to the development of Neuronavigation as a means
to define brain anatomy during surgery and to guide surgical interventions. The technology
was welcomed with much enthusiasm as it provided precise stereotactic definition of both the
brain anatomy and the boundaries of intracranial lesions. However, neuronavigation was
based on preoperatively acquired images and the brain shift caused by the surgical interven-
tion severely affected the accuracy and therefore the dependability of this technology.
Meanwhile several different technologies of intraoperative imaging were under development.
Ultrasonography (U/S), computed tomography (CT) and MRI are currently the most promi-
nent of these techniques. Initial designs were tested in the clinic and most were replaced by
never designs to accommodate clinical needs and to compensate for the shortcomings. The
financial burden of these sophisticated intraoperative imaging technologies was also a serious
consideration and had an important influence on equipment and facility designs. Intraoperative
imaging technology certainly did not stay confined to the field of neurooncology. Neurovas-
cular, pediatric, functional and spine surgery had different needs and these were fulfilled by
development of even more diversified technologies.
The increasing attention and interest on intraoperative imaging also necessitated interna-
tional interaction and collaboration and the Intraoperative Imaging Society was formed in
2007. The first Annual Meeting of the Intra-operative Imaging Society was held at the Hyatt
Regency Resort, Spa and Casino in Lake Tahoe-Nevada in 2008. After this very successful
meeting, the second meeting was held in Istanbul-Turkey from June 14 to 17, 2009. This book
brings together highlights from this second meeting of the Intraoperative Imaging Society. The
first section of the book gives an overview of the emergence and development of the
intraoperative imaging technology and it gives a glimpse on where the technology is heading.
Among all technologies, intraoperative MRI has received most of the attention due to
immense technical potential of this modality. Various new technologies have been developed
v
vi Preface
in the last decade and this led to very diverse designs. Therefore, we have divided this section
into parts discussing low, high and ultra-high field designs. The second, third and fourth
sections provide separate reports on each system. After reading these chapters the reader
should have a general idea on intraoperative MRI technology and know the pros and cons of
each design. The sections on CT and Ultrasonography are followed by a section with reports
from the most prominent centers which have attempted integrating different imaging technol-
ogies. The last one is a diverse section bringing together ancillary techniques as well as reports
on intraoperative robotic technology.
We believe that this book will provide an up-to date and comprehensive general overview of
the current intraoperative imaging technology as well as detailed discussions on individual
techniques and clinical results.
Istanbul, Turkey M.N. Pamir, T. Kırış

Frankfurt, Germany V. Seifert
March 2010
Contents
History- Development- Prospects of Intraoperative Imaging
From Vision to Reality: The Origins of Intraoperative MR Imaging . . . . . . . . . . . . . . . . . . 3

Black, P., Jolesz, F.A., and Medani, K.
Development of Intraoperative MRI: A Personal Journey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Fahlbusch, R.
Lows and Highs: 15 Years of Development in Intraoperative

Magnetic Resonance Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Schmidt, T., König, R., Hlavac, M., Antoniadis, G., and Wirtz, C.R.
Intraoperative Imaging in Neurosurgery: Where Will the Future Take Us? . . . . . . . . . . . 21

Jolesz, F.A.
Intraoperative MRI- Ultra Low Field Systems
Development and Design of Low Field Compact Intraoperative MRI

for Standard Operating Room . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Hadani, M.
Low Field Intraoperative MRI in Glioma Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

Seifert, V., Gasser, T., and Senft, C.
Intraoperative MRI (ioMRI) in the Setting of Awake Craniotomies

for Supratentorial Glioma Resection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Peruzzi, P., Puente, E., Bergese, S., and Chiocca, E.A.
Glioma Extent of Resection and Ultra-Low-Field ioMRI: Interim Analysis

of a Prospective Randomized Trial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Senft, C., Bink, A., Heckelmann, M., Gasser, T., and Seifert, V.
Impact of a Low-Field Intraoperative MRI on the Surgical Results

for High-Grade Gliomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Kırış, T. and Arıca, O.
Intraoperative MRI and Functional Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Gasser, T., Szelenyi, A., Senft, C., Muragaki, Y., Sandalcioglu, I.E.,
Sure, U., Nimsky, C., and Seifert, V.
vii
viii Contents
Information-Guided Surgical Management of Gliomas Using

Low-Field-Strength Intraoperative MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Muragaki, Y., Iseki, H., Maruyama, T., Tanaka, M., Shinohara, C.,
Suzuki, T., Yoshimitsu, K., Ikuta, S., Hayashi, M., Chernov, M., Hori, T.,
Okada, Y., and Takakura, K.
Implementation of the Ultra Low Field Intraoperative MRI PoleStar

N20 During Resection Control of Pituitary Adenomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Gerlach, R., Richard du Mesnil du Rochemont, Gasser, T.,
Marquardt, G., Imoehl, L., and Seifert, V.
Intraoperative MRI for Stereotactic Biopsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

Schulder, M. and Spiro, D.
The Evolution of ioMRI Utilization for Pediatric Neurosurgery:

A Single Center Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Moriarty, T.M. and Titsworth, W.L.
Intraoperative MRI - High Field Systems
Implementation and Preliminary Clinical Experience with the

Use of Ceiling Mounted Mobile High Field Intraoperative
Magnetic Resonance Imaging Between Two Operating Rooms . . . . . . . . . . . . . . . . . . . . . . . 97
Chicoine, M.R., Lim, C.C.H., Evans, J.A., Singla, A., Zipfel, G.J.,
Rich, K.M., Dowling, J.L., Leonard, J.R., Smyth, M.D.,
Santiago, P., Leuthardt, E.C., Limbrick, D.D., and Dacey, R.G.
High-Field ioMRI in Glioblastoma Surgery: Improvement

of Resection Radicality and Survival for the Patient? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Mehdorn, H.M., Schwartz, F., Dawirs, S., Hedderich, J.,
Dörner, L., and Nabavi, A.
Image Guided Aneurysm Surgery in a Brainsuite1 ioMRI Miyabi

1.5 T Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
König, R.W., Heinen, C.P.G., Antoniadis, G., Kapapa, T.,
Pedro, M.T., Gardill, A., Wirtz, C.R., Kretschmer, T., and Schmidt, T.
From Intraoperative Angiography to Advanced Intraoperative Imaging:

The Geneva Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Schaller, K., Kotowski, M., Pereira, V., Rüfenacht, D., and Bijlenga, P.
Intraoperative MRI - Ultra High Field Systems
Intraoperative Magnetic Resonance Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

Hall, W.A. and Truwit, C.L.
3 T ioMRI: The Istanbul Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

Pamir, M.N.
Intra-operative 3.0 T Magnetic Resonance Imaging Using a

Dual-Independent Room: Long-Term Evaluation of Time-Cost,
Problems, and Learning-Curve Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Martin, X.P., Vaz, G., Fomekong, E., Cosnard, G., and Raftopoulos, C.
Contents ix
Multifunctional Surgical Suite (MFSS) with 3.0 T ioMRI: 17 Months

of Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Beneš, V., Netuka, D., Kramář, F., Ostrý, S., and Belšán, T.
Intra-operative MRI at 3.0 Tesla: A Moveable Magnet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

Lang, M.J., Greer, A.D., and Sutherland, G.R.
One Year Experience with 3.0 T Intraoperative MRI in Pituitary Surgery . . . . . . . . . . 157
Netuka, D., Masopust, V., Belšán, T., Kramář, F., and Beneš, V.
Intraoperative CT and Radiography
Intraoperative Computed Tomography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

Tonn, J.C., Schichor, C., Schnell, O., Zausinger, S., Uhl, E.,
Morhard, D., and Reiser, M.
Intraoperative CT in Spine Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

Steudel, W.-I., Nabhan, A., and Shariat, K.
O-Arm Guided Balloon Kyphoplasty: Preliminary Experience

of 16 Consecutive Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Schils, F.
Intraoperative Ultrasonography
Intra-operative Imaging with 3D Ultrasound in Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . 181

Unsgård, G., Solheim, O., Lindseth, F., and Selbekk, T.
Intraoperative 3-Dimensional Ultrasound for Resection Control During Brain

Tumour Removal: Preliminary Results of a Prospective Randomized Study . . . . . . . . . 187
Rohde, V. and Coenen, V.A.
Advantages and Limitations of Intraoperative 3D Ultrasound

in Neurosurgery. Technical note . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Bozinov, O., Burkhardt, J.-K., Fischer, C.M., Kockro, R.A.,
Bernays, R.-L., and Bertalanffy, H.
Multimodality Integration
Integrated Intra-operative Room Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199

Ng, I.
Multimodal Navigation Integrated with Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

Nimsky, C., Kuhnt, D., Ganslandt, O., and Buchfelder, M.
Multimodality Imaging Suite: Neo-Futuristic Diagnostic Imaging Operating

Suite Marks a Significant Milestone for Innovation in Medical Technology . . . . . . . . . 215
Matsumae, M., Koizumi, J., Tsugu, A., Inoue, G.,
Nishiyama, J., Yoshiyama, M., Tominaga, J., and Atsumi, H.
Improving Patient Safety in the Intra-operative MRI Suite Using

an On-Duty Safety Nurse, Safety Manual and Checklist . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Matsumae, M., Nakajima, Y., Morikawa, E., Nishiyama, J.,
Atsumi, H., Tominaga, J., Tsugu, A., and Kenmochi, I.
x Contents
Operating Room Integration and Telehealth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223

Bucholz, R.D., Laycock, K.A., and McDurmont, L.
Other Intraoperative Imaging Technologies and Operative Robotics
Intra-operative Robotics: NeuroArm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231

Lang, M.J., Greer, A.D., and Sutherland, G.R.
Clinical Requirements and Possible Applications of Robot Assisted

Endoscopy in Skull Base and Sinus Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Eichhorn, K.W.G. and Bootz, F.
Robotic Technology in Spine Surgery: Current Applications

and Future Developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
Stüer, C., Ringel, F., Stoffel, M., Reinke, A., Behr, M., and Meyer, B.
Microscope Integrated Indocyanine Green Video-Angiography

in Cerebrovascular Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Dashti, R., Laakso, A., Niemelä, M., Porras, M., and Hernesniemi, J.
Application of Intraoperative Indocyanine Green Angiography

for CNS Tumors: Results on the First 100 Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Ferroli, P., Acerbi, F., Albanese, E., Tringali, G., Broggi, M.,
Franzini, A., and Broggi, G.
A Technical Description of the Brain Tumor Window Model:

An In Vivo Model for the Evaluation of Intraoperative Contrast Agents . . . . . . . . . . . . 259
Orringer, D.A., Chen, T., Huang, D.-L., Philbert, M., Kopelman, R.,
and Sagher, O.
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267

History- Development- Prospects of
Intraoperative Imaging
From Vision to Reality: The Origins of Intraoperative MR Imaging
Peter Black, Ferenc A. Jolesz, and Khalid Medani
Abstract Intraoperative MR imaging has become one of the MRT Magnetic Resonance Tomography
most important concepts in present day neurosurgery. The NIH National Institute of Health
brain shift problem with navigation, the need for assessment OR Operation Room
of the degree of resection and the need for detection of early
postoperative complications were the three most important
motives that drove the development of this technology. The Introduction
GE Signa System with the ‘‘double donut’’ design was the
world’s first intraoperative MRI. From 1995 to 2007 more The Beginning
than 1,000 neurosurgical cases were performed with the
system. The system was used by several different specialties
Intraoperative MR imaging has become one of the most
and in neurosurgery it was most useful for complete resec-
important concepts in present day neurosurgery. Many fac-
tion of low-grade gliomas, identification and resection of
tors led to the emergence of intraoperative MRI; the most
small or deep metastases or cavernomas, recurrent pituitary
important were:
adenomas, cystic tumors, biopsies in critical areas and sur-
gery in recurrent GBM cases. Main superiorities of the (1) Brain shift (movement of the brain relative to the cra-
system were the ability to scan without patient movement nium between the time of scanning and surgery). One of
to get image updates, the ability to do posterior fossa cases the main reasons for developing intraoperative imaging
and other difficult patient positioning, the easiness of opera- was the shift of cortical brain structures during surgery
tion using intravenous sedation anesthesia and the flexibility because of loss of CSF, shrinkage of brain tissue, and
of the system to be used as platform for new diagnostic and resection of the lesion. These changes make it difficult
therapeutic modalities. to navigate accurately with preoperatively acquired
images. This was particularly important for biopsies of
Keywords GE Signa Intraoperative MR MRI Navigation cystic lesions, for resection of deep lesions or those with
cysts or near the ventricles. Intraoperative MR can
Abbreviations accurately estimate changes to the brain which occur
during surgery, a property which is lacking in navigation
BWH Brigham and Women’s Hospital
systems using preoperatively acquired images [1 6].
CSF CerebroSpinal Fluid
(2) Assessing the degree of resection during surgery.
fMRI Functional MRI
Achieving greater tumor resection was another reason
GBM Glioblastoma Multiforme
for developing the intraoperative MR. Residual tumor
MEG Magnetoencephalography
can be detected using this modality and can be removed
MR(I) Magnetic Resonance(Imaging)
without a return to the operating room. This turned out
to be particularly important for pediatric tumors, glio-
P. Black (*) and K. Medani mas and pituitary adenomas.
Department of Neurosurgery, Brigham and Women’s Hospital, (3) Evaluating intraoperative complications. Intraopera-
Harvard Medical School, Boston, MA, USA
e mail: peterblackwfns@gmail.com
tive MR also has the capability to recognize acute
intraoperative changes such as hemorrhage, infarct,
F.A. Jolesz
Department of Radiology, Brigham and Women’s Hospital, Harvard and edema. These complications may be treated immedi-
Medical School, Boston, MA, USA ately, avoiding long-term disability.
M.N. Pamir et al. (eds.), Intraoperative Imaging, Acta Neurochirurgica Supplementum, Vol. 109, 3
DOI 10.1007/978 3 211 99651 5 1, # Springer Verlag/Wien 2011
4 P. Black et al.
The GE Signa System was the world’s first intraoperative The GE Signa MRT System
MRI. It used a novel ‘‘double donut’’ design in which the
magnetic field was between two open magnets (Fig. 1). The
In the GE 0.5 tesla Signa system, surgery is done directly in
main contributors to the system were Ferenc Jolesz and
the MRI scanner. All instruments must be non-ferromagnetic.
then Ron Kikinis and Richard Schwartz of the BWH
The device also acts as a powerful navigation system using
Dept of Radiology, Peter Black and Eben Alexander, III of
software called slicer [7, 8]. Neurosurgical MRT fellows
Neurosurgery, Marvin Fried from the Ear, Nose, and Throat
included: Tom Moriarty, Claudia Martin, Andrew Danks,
Department, Mory Blumenthal from General Electric, BWH
Kate Drummond, Vivek Mehta, Arya Nabavi, David Walker,
Biomedical Engineering, Maureen Hanley of nursing, Linda
Lorenzo Bello, Dennis Oh, Elizabeth Claus., Juan Ortega,
Aglio and George Topoulos of the anesthesia staff, and the
and Andrew Morokoff. They were instrumental in develop-
hospital administration.
ing the instrument and its applications.
An important feature the system was a collaborative
The slicer was a unique feature of our system developed
arrangement between the BWH and Children’s Hospital. Drs.
by the Surgical Planning Lab (SPL), an immensely talented
Mark Rockoff and Sulpiciano Soriano were particularly impor-
and powerful post-processing MR laboratory group led by
tant in this. On Wednesday each week the system became an
Ron Kikinis. With the slicer, the MRI itself became a navi-
outpost of the Children’s Hospital next door. The nursing and
gation system so no additional fiducials or registration was
anesthesia staff were all from Children’s; the patient was cared
needed.
for preoperatively and postoperatively at Children’s Hospital.
An example of a case operated using the GE Signa
As the years developed, Dr. Fried left the BWH and ENT
intraoperative MR is presented in Fig. 2: a 21-year-old
applications diminished. Radiation oncology under Dr.
woman came from Europe with intractable seizures of the
Anthony Amico became an important application, using
right leg and hand with speech arrest. The MRI showed a
the magnet for real time dosimetry for prostate cancer
tumor in the medial aspect of her left motor cortex (Fig. 2).
brachytherapy. For most of the life of the magnet, neurosur-
Surgery was successfully performed to resect her gangliocy-
gery and radiation oncology shared its use.
toma using the intraoperative MR. The images taken during
The magnet arrived at BWH in 1993. The components
the surgery show how the tumor was totally resected and the
included the magnet, console, anesthesia machines and
brain returned to normal (Fig. 3). The patient was discharged
monitors. The Midas REX drill system, operating micro-
from the hospital two days after surgery with no deficit. One
scope and CAVITRON ultrasonic device were developed
week after surgery, she was ready to go back to her home and
over the next few years with particular help from neurosur-
has been seizure-free for the four years following surgery.
gical fellows. In 1995, we performed the first brain tumor
Besides its real-time updates during surgery and its strong
craniotomy using the intraoperative MR [1].
navigation system, other advantages of the GE Signa system
included:
1. The joy of working with a great team.
2. The little risk of ferro-magnetic instrumental injury
because everything was screened before entering the
room.
3. The ability to scan without patient movement to get updates.
4. The ability to do posterior fossa cases and other position-
ing which is difficult for regular navigation systems.
5. The ready ability to operate on patients using intravenous
sedation anesthesia. For low grade gliomas, the combina-
tion of intraoperative navigation and intravenous sedation
anesthetic with brain mapping gave an accuracy and
safety not possible with other systems and ideal for the
patient.
6. The ability to use the scanner as a platform for new
diagnostic and therapeutic modalities such as scanning
with a patient upright to look at disc disease and use of
Fig. 1 The GE Signa intraoperative MRI laser hyperthermia for noninvasive ablation of a target.
From Vision to Reality: The Origins of Intraoperative MR Imaging 5
Fig. 2 21 year old female with a

gangliocytoma, before surgery
Fig. 3 21 year old female with a

gangliocytoma during surgery,
using the GE Signa intraoperative
MR. The lesion is localized with
the slicer in the coronal and
sagittal planes; it is removed
(lower left) and the bone is
replaced without difficulty. She
had no postoperative deficit
despite the proximity of the lesion
to primary sensory and motor
cortex
Fig. 4 Tumor types in 900 cases low grade

at BWH using the GE Signa gliomas
intraoperative MRI high grade
gliomas
recurrent
gliomas
other
Disadvantages of the system included: (cyst drainage, transsphenoidal procedures, etc). The major-
ity of tumors were low-grade gliomas (50%); high-grade
1. A modest restriction on positioning to get the ‘‘sweet
gliomas (24%), recurrent glioblastomas (26%), and other
spot’’ for imaging.
(10%) (Fig. 4).
2. Early on, limitation of instrumentation because of the
Forty percent of the low-grade gliomas were done with
requirement for non-ferromagnetic tools.
mapping and intravenous sedation anesthesia. In 38% of
3. Relatively low field strength compared with some other
cases, unexpected tumor was identified on intraoperative
systems.
repeat imaging. The Cincinnati, Erlangen, and Boston expe-
rience confirmed that 40 50% of cases showed residual
tumor that could be removed [9, 10]. The complication rate
Uses of Our System was only 5%, even in eloquent cortex, and the intraoperative
imaging predicted postoperative imaging very well.
More than 1,000 neurosurgical cases were done in the Signa In recurrent gliomas, intraoperative MR significantly
MRT system from 1995 to 2007. They included more than helped in resection of such tumors where the anatomical
800 craniotomies, 120 biopsies and 80 other procedures landmarks have been lost (Fig. 5). Intraoperative MR is
6 P. Black et al.
Fig. 5 Pre and post operative

recurrent glioblastoma, using the
GE Signa intraoperative MRI
Fig. 6 Pre and post operative

recurrent pituitary adenoma,
using the GE Signa intraoperative
MRI
also very useful in recurrent pituitary adenomas again demon- therapeutic approaches. These grants included program
strating residual tumor otherwise unresected [11] (Fig. 6). projects, national centers of excellence, RO-1’s, and foun-
We concluded that intraoperative MR was important for dation grants.
complete resection of low-grade gliomas, identification and
resection of small or deep metastases or cavernomas, recur-
rent pituitary adenomas, cystic tumors, biopsies in critical
areas and recurrent GBM [11 17]. The Future
In the near future, Intraoperative MR can be a broad base for

different advance imaging techniques such as fMRI, MEG,
Accomplishments MR Spectroscopy, as well as new therapeutic modalities
such as laser hyperthermia for lesion destruction, focused
Many improvements in patient care resulted from the use of ultrasonic surgery of tumors, robotic neurosurgery and, the
the intraoperative MRI at the BWH; they include: improve- new Advanced Multimodality Image Guided Operating
ment of the degree of resection of all tumors, shortening the (AMIGO) Suite at BWH [7, 18]. These topics will be cov-
duration of hospital stay, and reduction in the postoperative ered by Dr. Jolesz’s paper in this volume.
complication rate. Perhaps the most important task, however, is the creation
The unit was also extremely important in grant fund- of class-one evidence for the usefulness of intraoperative
ing, including large NIH grants for post-processing of imaging and the development of other applications including
images, applications of intraoperative imaging, and new spinal and functional neurosurgery.
From Vision to Reality: The Origins of Intraoperative MR Imaging 7
Conclusion 8. Hinks RS, Bronskill MJ, Kucharczyk W, Bernstein M, Collick BD,

Henkelman RM (1998) MR systems for image guided therapy.
J Magn Reson Imaging 8:19 25
The early days of intraoperative MR imaging were pioneer- 9. Darakchiev BJ, Tew JMJ, Bohinski RJ, Warnick RE (2005)
ing, exciting, and visionary. Particularly important were the Adaptation of a standard low field (0.3 T) system to the operating
teams developed, the concept of the power of this modality, room: focus on pituitary adenomas. Neurosurg Clin N Am
and the stimulation of the next generation to move the field 16:155 164
10. von Keller B, Nimsky C, Ganslandt O, Fahlbusch R (2004) Intrao
forward. The creation of the intraoperative imaging society perative MRI in 62 patients with pituitary adenoma. Deutsche
is an important step in consolidating this field and it will be Gesellschaft für Neurochirurgie. Ungarische Gesellschaft für
exciting to see how this develops. Neurochirurgie. 55. Jahrestagung der Deutschen Gesellschaft für
Neurochirurgie e.V. (DGNC), 1. Joint Meeting mit der Ungar
Conflict of interest statement Dr. Jolesz has received consultant ischen Gesellschaft für Neurochirurgie. Köln, 25. 28.04.2004.
fees from General Electric, who pioneered this system. Düsseldorf, Köln: German Medical Science DocMO.01.05
11. Martin CH, Schwartz R, Jolesz F, Black PM (1999) Transsphenoi
dal resection of pituitary adenomas in an intraoperative MRI unit.
Pituitary 2:155 162
References 12. Lacroix M, Abi Said D, Fourney DR, Gokaslan ZL, Shi W,
DeMonte F, Lang FF, McCutcheon IE, Hassenbusch SJ, Holland E,
Hess K, Michael C, Miller D, Sawaya R (2001) A multivariate
1. Black PM, Moriarty T, Alexander ER, Stieg P, Woodard EJ, analysis of 416 patients with glioblastoma multiforme: prognosis,
Gleason PL, Martin CH, Kikinis R, Schwartz RB, Jolesz FA extent of resection, and survival. J Neurosurg 95:190 198
(1997) Development and implementation of intraoperative mag 13. Lagerwaard FJ, Levendag PC, Nowak PJ, Eijkenboom WM,
netic resonance imaging and its neurosurgical applications. Neuro Hanssens PE, Schmitz PI (1999) Identification of prognostic
surgery 41:831 842, discussion 842 845 factors in patients with brain metastases: a review of 1292 patients.
2. Hata N, Nabavi A, Wells WMR, Warfield SK, Kikinis R, Black Int J Radiat Oncol Biol Phys 43:795 803
PM, Jolesz FA (2000) Three dimensional optical flow method for 14. Moriarty TM, Quinones Hinojosa A, Larson PS, Alexander ER,
measurement of volumetric brain deformation from intraoperative Gleason PL, Schwartz RB, Jolesz FA, Black PM (2000) Frameless
MR images. J Comput Assist Tomogr 24:531 538 stereotactic neurosurgery using intraoperative magnetic resonance
3. Hill DL, Maurer CRJ, Maciunas RJ, Barwise JA, Fitzpatrick JM, imaging: stereotactic brain biopsy. Neurosurgery 47:1138 1145,
Wang MY (1998) Measurement of intraoperative brain surface discussion 1145 1146
deformation under a craniotomy. Neurosurgery 43:514 526, dis 15. Pollack IF, Claassen D, al Shboul Q, Janosky JE, Deutsch M
cussion 527 528 (1995) Low grade gliomas of the cerebral hemispheres in children:
4. Nabavi A, Black PM, Gering DT, Westin CF, Mehta V, Pergolizzi an analysis of 71 cases. J Neurosurg 82:536 547
RSJ, Ferrant M, Warfield SK, Hata N, Schwartz RB, Wells WMR, 16. Schwartz RB, Hsu L, Black PM, Alexander ER, Wong TZ,
Kikinis R, Jolesz FA (2001) Serial intraoperative magnetic reso Klufas RA, Moriarty T, Martin C, Isbister HG, Cahill CD,
nance imaging of brain shift. Neurosurgery 48:787 797, discussion Spaulding SA, Kanan AR, Jolesz FA (1998) Evaluation of
797 798 intracranial cysts by intraoperative MR. J Magn Reson Imaging
5. Nimsky C, Ganslandt O, Hastreiter P, Fahlbusch R (2001) Intrao 8:807 813
perative compensation for brain shift. Surg Neurol 56:357 364, 17. Schwartz RB, Hsu L, Wong TZ, Kacher DF, Zamani AA,
discussion 364 365 Black PM, Alexander ER, Stieg PE, Moriarty TM, Martin CA,
6. Reinges MH, Nguyen HH, Krings T, Hutter BO, Rohde V, Gilsbach Kikinis R, Jolesz FA (1999) Intraoperative MR imaging guidance
JM (2004) Course of brain shift during microsurgical resection of for intracranial neurosurgery: experience with the first 200 cases.
supratentorial cerebral lesions: limits of conventional neuronaviga Radiology 211:477 488
tion. Acta Neurochir (Wien) 146:369 377, discussion 377 18. Hynynen K, Vykhodtseva NI, Chung AH, Sorrentino V, Colucci V,
7. Black PM (2003) Current and future developments in intraopera Jolesz FA (1997) Thermal effects of focused ultrasound on
tive imaging: MRI. In: Apuzzo MLJ (ed) The operating room for the brain: determination with MR imaging. Radiology 204:
the 21st century. AANS, Park Ridge, pp 45 51 247 253
Development of Intraoperative MRI: A Personal Journey
Rudolf Fahlbusch
Abstract The initial attempts at intraoperative image guid- understand immediately, which machine Peter was going to
ance and imaging dates back to early 1980s. Since then introduce to us there. Some days later, in his computer-
Neuronavigation and intraoperative imaging technologies laboratory, which came out to be the first one in neurosur-
were developed in parallel. This works aims at summarizing gery up to my knowledge, he presented to me the first
the developments and giving an insider’s view into the prototype of a pointer related neuro-navigation system.
beginning stage of these technologies. The successes and When I asked him to provide me with the first commercial
obstacles encountered in the first few decades are relayed navigation system in Europe, this happened to us in Erlangen
from the angle of one of the initial developers. in 1993. Peter Heilbrunn’s pilot-system (‘‘machine vision’’)
was later commercialized as the Stealth System by Surgical
Keywords Intraoperative MR MRI Navigation Navigation company, directed by Kurd Smith, who helped
intensively and frequently to integrate the system in our OR.
Dreams in my neurosurgical life were focussed on continous It was Richard Buchholz from St. Louis, who completed the
improvement of surgical results and I was convinced that device, introducing also special LEDS (This navigation sys-
this could be achived by early or real time testing. In the tem was distributed later by Sofamor-Danek and latest by
seventies I introduced endocrinological methods to predict Medtronics). It was at a much earlier opportunity in the
early outcome of pituitary surgery. In the eighties neuro- middle of the seventies when I would have had the chance
physiological monitoring for brainstem- and cerebellopon- to realize the significance of an early mechanical navigation
tine angle (CPA) tumors improved functional surgical results. system, I overlooked it a long while, after its principles were
In the nineties computer and engineering sciences were presented to us by Eiju Watanabe, when he was a research
incorporated in surgical planning and surgical manoeuvres, fellow in Erlangen, coming from Tokyo University. Retro-
in order to obtain a safer and more accurate brain tumor spectively I was not completely convinced about the practi-
surgery. From the very beginning neuronavigation and cal use of this-at that time not so accurate system.
intraoperative MRI were from my point of view two Nevertheless we used it for a while in traumatology cases.
parallel developments, supporting each other. This shall be Watanabe’s cooperation with his teacher Kintomo Takakura
illustrated by my personal experience. Certainly this is not a was published in 1987 [1]. Today he is regarded as the
systematic complete presentation of all efforts in this field. ‘‘father’’ of the modern neuro-navigator.
There are moments for decisions in our professional lives, Other pioneers in this field of intraoperative imaging
which can be regarded as destiny, maybe a favourable were Patrick Kelly with ‘‘volumetric stereotaxy’’ in 1979
opportunity, prepared already in our inner development. [2], Schlöndorff, a German ENT-professor with ‘‘computer
Such an event happened to me in January 1992, when assisted surgery’’ in 1986 [3] and Alim-Louis Benabid, who
Peter Heilbrunn, at that time Chairman of the Department constructed a stereotactic robot for performing biopsies and
of Neurosurgery University of Utah in Salt Lake City, and positioning of deep seated electrodes in 1987 [4]. Together
main organiser of the Lende Winter Meeting, invited me with Christian Saint Rose he had also developed a neurona-
to his winter house in Snowbird. When he asked Robert vigation system (see below). With the beginning of the
Spetzler and myself for a visit of his bedroom we could not nineties the former neurosurgeon and later radiologist
Frank Jolesz founded together with the radiologist and com-
puter scientist Ron Kikinis the first Surgical Planning Lab
R. Fahlbusch
International Neuroscience Institute Hannover, Hannover, Germany in Boston. In 1995 Kazuhiru Hongo introduced navigated
e mail: fahlbusch@ini hannover.de micromanipulation onto the way of robotics. As early as
10 R. Fahlbusch
1991 Dade Lundsford introduced intraoperative CT (ioCT) ogy, positioned close by. A narrow working area for the
in Pittsburgh, but gave it up due to minor resolution for neurosurgeon was the price for receiving online MRI data.
imaging of brain tumors. The neurosurgeons could use this permanent image informa-
The following experiences shall illustrate how close tion during their resection of a brain tumor, which allowed
developments in neuronavigation and intraoperative MRI them to follow and compensate the brain shift for navigation
were running parallel and, supported each other. In the for the first time. I was fascinated too by the Surgiscope, a
beginning of the nineties a technical engineer from Zeiss highly sophisticated navigation system, developed originally
company, Mr Marcovic, stayed with us in our operating by Alim-Louis Benabid and Christian Saint Rose, pioneers
room in Erlangen as a guest (observer) After some days he in neuronavigation, Saint Rose had demonstrated convinc-
asked me, how far I would be interested to see the MR- ingly its accuracy to me during a resection of a pediatric
images no longer in the traditional way on the screen at the glioma in Paris. Later it came out that Electa company, the
wall, but within the eyepieces of the microscope. I was fully distributer of the Surgiscope, and the distributor of the first
convinced by this principle, when Mr. Marcovic and Mr. open low field MRI (Magnetom) Siemens had no common
Luber demonstrated me the first pilot microscope in the ‘‘Schnittstelle’’ this was for me the end of a potential
Zeiss laboratories later in Tuttlingen, which could offer realisation.
projection of MR images into the eyepiece of the microscope Our wish to realise an intraperative MRI system together
for the use of navigation. Furthermore I could discover, that with navigation in Erlangen was favoured by the fact, that the
this Zeiss MKM was not only a tool for neuro-navigation, former director of Siemens Medical Solution, Dr. Grassman
but offered also robotic potential. Its movement from the became Director of Zeiss, Tuttlingen and that he was followed
stand-by position to the point of view position in the by Prof Reinhardt, with whom we cooperated before with
operating field could be ordered by voice. In my eyes this some projects of MRI visualisation of pituitary tumors and
was the birthday of microscope-guided navigation. So far we surrounding arteries (MR angiography.)
were working with pointer-guided systems for example with One year after our common decision, induced by
the Stealth navigation system. Some months later I visited Mr Schöck, the chancellor of Erlangen University, our new
the Siemens development laboratories in Erlangen and OR suite with the 0.2 T open MRI and the Zeiss MKM
became aware of a newly developed open MRI. The 0.2 T (Figs. 1 and 2), could be realised together with the neuro-
machine offered not only an acceptable resolution for diag- surgeon Ralf Steinmeier. We were able to perform our first
nosis, but could document manoeuvers in orthopaedic sur- operation in March 1996-accompanied by a lot of worries.
gery, in a way of combined imaging and navigation, showing Would everything run really well during the transport of our
nearly on line movements of instruments. patient with an open trepanation after a brain tumor resection
It was during the Meeting of the ‘‘International Pituitary from his position on the operating table, into the gantry of
Neurosurgeons Society’’ in Bamberg, at the opportunity of a the MR scanner, then docking this table to the MR machine,
social evening event in a beer cellar, when I asked my colleagues, which took this over as examination table-and the same
if they have heard also that someone in United States is going procedure backwards? What about sterility during this
to introduce MRI in the operating room. I was so much sur- prolonged surgery? How would the images look like,
prised that my table neighbour Peter Black said, ‘‘Yes, it is me’’. would the resolution of images be sufficient, how intensive
It was only a question of time when I went to see the first would artefacts influence the results? For pituitary surgery
equipment for ioMRI, the 0.5 T Signa SP (Double donut), in
Boston. This was a development of GE in cooperation and
on demand of Frank Jolesz and Peter Black, probably as a iopMRI:
result of their experiences they had gained in their Surgical Version 1a
Navigation Planning Lab: the problem of brain shift. After
an intensive preparation time of more then 2 years-including
safety aspects for patients and medical staff being in or close
to a magnetic field for a longer time- the first operation a
brain biopsy was performed in June 1995-the first trepana- 0,2T MRI
Stealth Navi
tion was performed in January 1996.Peter Black had invited
me to demonstrate a biopsy for a brain tumor in fall 1995, I Zeiss MKM
could observe ENT doctors using a copper-endoscope for
surgery of the paranasal sinuses. It was obvious that this
Fig. 1 The original, initial ioMRI (open Magnetom (version 1a) in
continuously running magnetic field tolerated within its Erlangen. Above: view from the room with the Magnetom Open into the
field strength only MR compatible equipment, starting with OR room. Below: view from the OR with Zeiss MKM and Stealth
surgical instruments ending with machines for anaesthesiol- station to the MRI room
Development of Intraoperative MRI: A Personal Journey 11
Fig. 2 Docking manoeuvre of

the OR table with the MR scanner
(version 1a)
we learned to avoid drilling artefacts by using porcelain

coated drills instead of stainless steel drills and to insert a
small wax plate at the sellar floor after tumor resection to
separate the intrasellar space from the sphenoid sinus, with
its bleeding artefacts during data acquisition.
From the very beginning of developing intraoperative
MRI there existed different options and concepts for its
realisation, which could not be tested in an experimental
way before:
A continuous magnetic field for on line imaging with
surgery within the magnetic field. iopMRI version1b :compatible Navi-micr. NC4
A magnet separate from the operating field, where the
advantages of microscope navigation could be used. Another Fig. 3 ioMRI (version 1b) integration of the compatible navigation
concept decision was related to the field strength: low field microscope Zeiss NC4 close to the modified diagnostic and therapeutic
operation table
vs high field: This concept as well as the concept of the
relation of the patient’s to the magnet position is till today in
discussion: Shall the patient be transported into the gantry of compatible coils as well as a special operating table. Where
the magnet or the magnet to the patient. We used pointer their operating room was large and the MRI room small it
related navigation and microscope guided navigation as was the other way around in our concept. This was an
well. At the beginning it was necessary to have two different unforeseen advantage, since we could work later on with
rooms, one room for the surgical procedure, including navi- the newly introduced Zeiss NC4 closer to the MR gantry,
gation and the other room for MR control. Both were within the room with the Magnet, some tests have demon-
connected to each other, but could be separated by a shielded strated before compatibility of the equipment (Fig. 3, version
door. This had the advantage that the MR room could be 1b). From now on we could operate on the diagnostic MRI
used during surgery also for examinations of other patients table which was connected with a special compatible head
(a concept in ioMRI systems, which is still used today for holding device, the time for major transportation seemed to
commercial reasons.) At that time we could not perform be passed away. We also found out that we could work
surgery and MRI examinations in one common room. without any compatibility problems outside and at the so
Tests have demonstrated that there was no compatibility of called 5 Gauss line (Fig. 4).
equipment. 1995 and 1996 in parallel the Heidelberg group Chronologically the introduction of a high field strength
of neurosurgeons (Stefan Kunze, Christian Wirtz, Volker MRI, the 1.5 T(Philips) was just following the developments
Tronnier) had developed ‘‘our’’ concept too, they introduced of low field MRI systems in Boston, Erlangen and Heidelberg
12 R. Fahlbusch
cussion with the more convincing images, gained by high

field systems.
The real breakthrough in the use of intraoperative MRI
happened from our point of view, when we were able to
combine functional navigation with high field 1.5 T MRI in
Erlangen (Fig. 5 and 6). After intensive planning and devel-
opments together with Christopher Nimsky and the industri-
al companies we could integrate the Siemens 1.5.T MRI
Sonata and the BrainLab Vecor Vision sky system naviga-
tion (Vilsmaier, Ehrke, Kraft) in our new concept. This
included also an integrated head coil for automatic registra-
tion intraoperatively. Functional MRI, using tools of MEG
(Kober, Grummich) and the Bold effect of MRI (together
with Oliver Ganslandt) allowed accurate localisation of elo-
quent areas such as sensor, motor, speech area (Broca and
Wernicke). Functional and morphological data could be
segmented, used for intraoperative navigation and could be
upgraded after intraoperative MRI control. With the version
of a rotating table the patient’s transport was solved, we
Fig. 4 Head position and surgery at the s.c 5 Gauss line, the safety could work at the 5 Gauss line with normal instrumentation.
border for safe surgery (version 1b) Even epilepsy surgery with EEG and ECG detections was
successfully performed (Michael Buchfelder, Johann Rom-
of in 1997. The radiologist Charles Truwitt, together with the stöck). An experimental test documented that even the use of
neurosurgeon Walter Hall gained their first experiences with a robotic system close to the operating able was also toler-
biopsies ,later with trepanations at the University of Minnea- ated. The advantages of high field strength for quicker ac-
polis. However they were unable to introduce navigation at quisition time and improved anatomical, imaging became
the beginning. Initially biopsies were taken close to the obvious: the intraoperative resolution quality was the same,
gantry of the MRI on the specially equipped diagnostic even superior, to the preop one, we came to the statement
table, later on the trepanations had to be performed depart that there is a necessity to incorporate functional imaging
from the gantry and the table had to be transported some and visualisation into Ors, since about 30 40% of resectable
meters from a more distant place for surgery. In contrast to brain tumors (pituitary adenomas and gliomas for example),
this concept of patient to the magnet Garnette Sutherland are overlooked or not visible initially and can be resected
developed a system for magnet to the patient (1.5 T MRI, safely, which can be documented immediately. The other
IMRIS) in Calgary, Canada in 1996. high field strength advantages are functional imaging which
In contrast to these efforts of early installation of allowed us also to complement cortical mapping with DTI-
high field magnets the development of low field system tractography (Christopher Nimsky) and to introduce Proton
continued, stimulated by the attraction of lower costs and spectroscopy with Ganslandt and the physicists Moser and
earlier availability. On one hand John Koivokangas installed Stadlbauer from Vienna a potential for higher cytoreduction.
an open 0.36 T MRI (Philips) in Oulu, Finland in 1996/1997, Meanwhile it became obvious that there was no need for
following the concept patient to the magnet. A similar de- more then one, maximal three intraoperative MRI controls,
vice was introduced by Ronald E. Warnick and John Tew which made the permanent online detecting of date unnec-
with the 0.3 T MRI, produced by Hitachi company and was essary. The first ioMRI system, the double doughnut, was no
later used by Kintomo Takakura and Tomokatsu Hori in longer promoted by its industrial company. With the avail-
Tokio. The first an ultra low field MRI system in a magnet ability of 3 T MRI for diagnostic purpose, with advantages
to patient system was developed in Israel and introduced to for functional and metabolic imaging, there introduction in
patients by Moshe Hadani in Tel Aviv and by Peter Carmel the OR was only a question of time. In 2005 and 2006
and Michael Schulder in Newark: the Odin Pole Star N10 Necmettin Pamir (Siemens), Christian Raftopoulos (Philips)
had a field strength of 0.12 T in 2000, the N15 version a field and Robert Spetzler (GE) started to installed these systems
strength of 0.15 T.The transportable magnet is positioned for therapeutic use.
below the operating ,when not needed and can be swinged In 2007 we opened the INI-BRAIN SUITE at the Interna-
upwards to the head, when intraoperative maneuvers shall be tional Neuroscience Institute: the first Open (70 cm bore!)
controlled. Till today the discussion about the value of a intraoperative high field MRI for therapeutic use (Fig. 7).
‘‘useable resolution’’ of the low field images, stands in dis- For the first time I was convinced to plan ‘‘2 operating
theatres in one room’’ simultaneously. The actual one for Intraoperative MRI: Gimmick or Godsend. Meanwhile
stable daily use on high quality level, the later one a 3 T, in about 150 ioMRI systems are installed world wide, domi-
case that the actual problems of 3 T MR will be solved: nated by low field systems. All systems have different
surface distortion, workflow with table transportation. Inte- advantages and disadvantages for:
gration of ceiling mounted microscope Zeiss Pentero. Retro- 1-image/quality benefit, 2-acquisition time, 3-imaging
spectively the time from planning and decision for an MRI modalities workflow, and 4-costs
lasted 1 year for the 0.2 T version, nearly 3.5 year for the The scientific development of intraoperative imaging
1.5 T version in Erlangen and less then 2 year for the 1.5 were accompanied by a number of meetings on local and
version in Hannover. The walls of the OR are isolated in a international level. Since 1996/1997-on the US level Peter
way required for 3 T systems. The exchange from a 1.5 T to a Black ad GE initiated symposia and workshops with GE
new 3 T (Siemens Verio) would prolong the 5 Gauss line only users as well as guests using other equipments in Boston
approximately 20 cm (4.5 4.2 m distance) from the Gantry. and at the opportunity of Congresses of AANS, CNS and
It happened in October 2004 in Dresden, at the opportu- WFNS. In Europe symposia were organised organised by
nity of a joint meeting of the German Society and German René L. Bernays (the later first president of the Intraopera-
Academy with the American Academy of Neurosurgeons tive Imaging Society in Foundation) in Zürich, Switzerland
that the necessity of ioMRI was accepted obviously in the and by Johannes Schramm (the later EANS president) in
international neurosurgical community. This was the definite Seeheim Jugenheim, Germany as an EANS Symposium. In
result of a contemporary formal discussion between Spetzler Germany the first symposium for navigation and intraopera-
and Fahlbusch in the topic Controversies in Neurosurgery: tive MRI was organised by Joachim Gilsbach, Aachen, fol-
Fig. 5 (a) Concept of the

interdisciplinary Neurocenter
(1996) Intraoperative MRI and
Coop with
functional Neuronavigation the
Physicist,
Erlangen Concept to centralise
Computer scientists,
neurodata for fusion, post
Industry,
processing and finally surgery.
Precondition
(b) View into the For research/grants
interdisciplinary Neurocenter, DFG;SFB
opened 7/2000
14 R. Fahlbusch
Fig. 6 IoMRI version 2: 1.5 T

MRI Siemens Sonata with a Erlangen II
rotating table and the Zeiss NC4 Intraoperative high-field 1.5T MRI Siemens Sonata : 4/02-1/06
(later Pentero) BrainLab Vector
Sky Navigation system in one
room. First operation in February
2002
Lesion Type Number of Patients

Glioma 233
Pituitary Adenoma 203
Craniopharyngioma 27
Cranio Cyst Puncture 25
Non-Lesional Epilepsy 66
Misc. Brain Tumors 132
April 2002 - January 2006 686
In 30–40 % Impact on Lesion Resection or Surgical Strategy after iop MRI
a b
INI Brain SUITE
13.2.2007 – 5.6.2009 243 iop procedures
PACS
Data
NS-Workstation
(iPlan) 1. Floor
c
Iop MRI +fNavig
OR 1
Stereotaxy Suite WORKSTATION
(calibrated bp X-ray) OR 2
“OR1 - System“ OR 3
Online + Archive
OR 4OR 5 OR 6
Angiography Suite
PACS
5th Floor - Operating Rooms NS Workstation

1 Floor
Fig. 7 (a) INI BrainSuite (version 3) Installation procedure of the MRI mounted Pentero. (c) Localisation of the BrainSuite within the OR tract
scanner at the INI Hannover. (b) View into INI BrainSuite. First on fifth floor. Structural engineering for construction stability was
operation 13.2.2007 Open 1.5 T MRI Siemens Espree with 70 cm planned already with the construction of the INI building, finished in
Gantry with BrainLab Navigation and Zeiss NC4, later with a ceiling 7/2000
lowed by Rudolf Fahlbusch and Christopher Nimsky in The First Society for Computer and Robotic Assisted
Erlangen, Maximillian Mehdorn and Arya Nabavi in Kiel. Surgery was founded in Leipzig, Germany in 2001. It was
At this time it became obvious that neurosurgeons had to expected that this scientific community would support fur-
cooperate and incorporate with engineering and computer ther development. Later on members of industrial companies
scientists. In 1996 we planned the ‘‘Neurocenter’’ in the organised user meetings. BrainLab 1007 in Houston and
Head. Clinic Erlangen, which we could open in July 2000, 2008 in Singapore, whereas Medtronic organised a previous
were experts and users of all clinical neurodisciplines and meeting for the foundation of the International Imaging
computer scientists gathered neurodata for post-processing Society in Lake Tahoe in 2008, under the guidance of
procedures, which could be used for patients’ operation.(Led Moshe Hadani, Rene Bernays and Michael Schulder. In
by the physicists Kober and Grummich, the computer scien- June 11 14 2009 the Intraoperative Imaging Society was
tist Hastreiter). Without this symbiotic input, research fund- definitely founded in Istanbul, where Necmettin Pamir, one
ing would not have been successful (Fig. 5a, b). of the initial pioneers of intraoperative 3 T-MRI, hosted the
A round table with participants from clinical users such first official conference of this society.
as neurosurgeons and radiologists, from industrial com-
panies, university administration and ministeries of the Conflict of interest statement We declare that we have no conflict
government was established to work out concepts for of interest.
MEDICAL PROCESS OPTIMIZING, this included also
economic aspects. Generally open problems in image Acknowledgement I especially would like to thank Christopher
Nimsky for his fundamental work in this field, he as well as other
guided surgery are the reduction and concentration of quoted authors are contributors of this book.
image data, application accuracy, visualization from 2D
to 3D improved, as well as model based safety corridors
for surgical maneuvers. Further perspectives for intra-
operative MRI scanning are the following: less cost References
intensive systems should be provided, and improvement
of ergonomic integration into the OR environment and 1. Watanabe E, Watanabe T, Manaka S, Mayanagi Y, Takakura K
workflow a vision would be a nearly invisible and nearly (1987) Three dimensional digitizer (neuronavigator): new equip
online flat or tabletop magnet-integration of therapeutic ment for computed tomography guided stereotaxic surgery. Surg
devices (e.g. smart intelligent instruments, thermoablation, Neurol 27:543 547
2. Kelly PJ, Kall BA, Goerss S (1984) Transposition of volumetric
focused ultrasound). information derived from computed tomography scanning into
Within the next 10 years we can expect the following stereotactic space. Surg Neurol 21:465 471
developments in MR technology: ‘‘a widespread shift 3. Schlondorff G, Mosges R, Meyer Ebrecht D, Krybus W, Adams L
to higher field systems (3 T), further improvement of coil (1989) CAS (computer assisted surgery). A new procedure in head
and neck surgery. HNO 37:187 190
technology, including further increase of channels, introduc- 4. Benabid AL, Cinquin P, Lavalle S, Le Bas JF, Demongeot J, de
tion of molecular MRI agents and combined modality Rougemont J (1987) Computer driven robot for stereotactic sur
methods’’ [5]. gery connected to CT scan and magnetic resonance imaging.
Meanwhile hybrid systems for io-imaging are installed, Technological design and preliminary results. Appl Neurophysiol
50:153 154
respectively in development: Mitsumori Matsumae, Tokai 5. Blamire AM (2008) The technology of MRI the next 10 years? Br
university, Japan, connected the OR to a CT and MR suite as J Radiol 81:601 617
well [6]. In Boston the planned AMIGO suite will connect 6. Matsumae M, Koizumi J, Fukuyama H, Ishizaka H, Mizokami Y,
the OR with 3 T MRi and a PET-CT. Meanwhile a com- Baba T, Atsumi H, Tsugu A, Oda S, Tanaka Y, Osada T, Imai M,
Ishiguro T, Yamamoto M, Tominaga J, Shimoda M, Imai Y (2007)
mercial PET-MR is available (Werner Siemens Foundation, World’s first magnetic resonance imaging/X ray/operating room
Radiology Tübingen, Germany) and will open further suite: a significant milestone in the improvement of neurosurgical
aspects of application. diagnosis and treatment. J Neurosurg 107:266 273
The escalating resources in intraoperative imaging in-
clude also other imaging developments, the already estab-
lished ultrasound technology and the newly developing
optical imaging, such as optical coherence tomography and Selected Readings
multiphoton excitation microscopy-working on a cellular
level, which for itself or in combination with ioMRI under- 7. Black PM, Moriarty T, Alexander E 3rd, Stieg P, Woodard EJ,
line the scientific efforts, which will allow the increasing Gleason PL, Martin CH, Kikinis R, Schwartz RB, Jolesz FA (1997)
Development and implementation of intraoperative magnetic reso
establishment of this kind of adaptive tumor resection as a
nance imaging and its neurosurgical application. Neurosurgery
standard procedure. 41:831 842
16 R. Fahlbusch
8. Blamire AM (2008) The technology of MRI: the next 10 years. Br J magnetic resonance imaging with the magnetom open scanner
Radiol 89:601 617 concepts, neurosurgical indications, and procedures: a preliminary
9. Buchholz R, MacNeil W, McDumont L (2004) The operating room report. Neurosurgery 43:739 747
of the future. Clin Neurosurg 51:228 237 12. Watanabe E, Watanabe T, Manaka S, Mayanagi Y, Takakura K
10. Fahlbusch R, Samii A (2007) A review of cranial imaging techni (1987) Three dimensional Digitizer (Neuronavigator): new equip
ques: potential and limitations. Clin Neurosurg 54:100 104 ment for computed tomography guided stereotactic surgery. Surg
11. Steinmeier R, Fahlbusch R, Ganslandt O, Nimsky C, Buchfelder M, Neurol 27:543 547
Kaus M, Heigl T, Lenz G, Kuth R, Huk W (1998) Intraoperative
Lows and Highs: 15 Years of Development in Intraoperative
Magnetic Resonance Imaging
T. Schmidt, R. König, M. Hlavac, G. Antoniadis, and C.R. Wirtz
Abstract Intraoperative magnetic resonance imaging time and improved postoperative performance after gross
(ioMRI) during neurosurgical procedures was first imple- total resection [5 9]. Additionally in early postoperative
mented in 1995. In the following decade ioMRI and image MRI residual tumor could be detected particularly in supra-
guided surgery has evolved from an experimental stage into a tentorial glioma in up to 70% of operations [5]. These results
safe and routinely clinically applied technique. The develop- reveal an enormous discrepancy between the surgeons per-
ment of ioMRI has led to a variety of differently designed ception of resection extent and objective radicality. In this
systems which can be basically classified in one- or two-room respect the implementation of neuronavigation has been
concepts and low- and high-field installations. Nowadays associated with high expectations in regard to the increase
ioMRI allows neurosurgeons not only to increase the extent of radicality. These expectations have been disappointing at
of tumor resection and to preserve eloquent areas or white least in part, presumably due to the intraoperative brain shift
matter tracts but it also provides physiological and biological [10]. Consistently this data and the high sensitivity of MRI in
data of the brain and tumor tissue. This article tries to give a detecting brain tumors generated a necessity that led to the
comprehensive review of the milestones in the development integration of ioMRI. The ioMRI suite development has
of ioMRI and neuronavigation over the last 15 years and followed two general concepts. The first concept developed
describes the personal experience in intraoperative low and by the Black group is based on continuously refreshed
high-field MRI. images [1, 11] and was used to perform magnetic resonance
imaging (MRI) during surgery using a dedicated MRI suite
Keywords Image-guided neurosurgery Intraoperative in which surgery took place within the magnet at a midfield
magnetic resonance imaging Low field high field MRI strength of 0.5 T. It allowed the surgeon access to the patient
Neuronavigation during the imaging procedure through the 56-cm-wide gap
of the open-bore and provided the surgeon with nearly real-
time neuronavigation. The second alternative concept was to
separate the place of surgery and imaging into a one- or two-
Introduction room suite design. ioMRI was performed discontinuously
during the surgical procedure by either transferring the pa-
tient to the MRI scanner or move the MRI to the patient
The concept of image guided surgery has resulted in a
[4, 12 17]. Although surgery has to be interrupted for imag-
strategic shift in MRI from diagnosis to treatment. Since the
ing this suite design allows the neurosurgeon to use conven-
fundamental neurosurgical studies at several institutions [1 4]
tional surgical tools and therefore reduces the fringe field
in the 1990s, intra-operative magnetic resonance imaging
related restrictions. In the following decade the tremendous
(ioMRI) and neuronavigation now enables neurosurgeons to
improvements in MRI technique directly found their way
increase radicality of tumor resection while preserving func-
into the neurosurgical operation room. At present the
tionality. Numerous past and contemporary studies could
ioMRI systems can be divided into low- or ultra-low-field
demonstrate a statistically significant prolongation of survival
installations [18 20] and high-field installations at 1.5 T or
3 T [12, 13, 15, 17, 21]. At least advanced imaging modal-
ities such as diffusion tensor imaging and tractography
T. Schmidt (*), R. König, M. Hlavac, G. Antoniadis, and C. Wirtz
could be successfully integrated to the microscope-based
Department of Neurosurgery, District Hospital Günzburg, University of
Ulm, Ludwig Heilmeyer Straße 2, 89312 Günzburg, Germany neuronavigation both for preoperative and intraoperative
e mail: thomas.e.schmidt@uni ulm.de surgical planning [22 24]. However, cost constrains and
18 T. Schmidt et al.
the lack of studies that prove the superiority of high-field- of surgically induced contrast enhancement representing a
systems in terms of progression-free-survival and longterm- potential risk of misinterpretation of ioMRI [28]. Surgically
survival have to be considered. Against the background of diminished contrast enhancement in high-grade residual
15 year of intraoperative magnetic resonance imaging and tumor that showed a strong contrast enhancement in the
neuronavigation this articles reviews the personal clinical preoperative dataset could also be observed in some cases.
and scientific experience at low-field and high-field ioMRI. For that reason consideration of the location, configuration
and time course of contrast enhancement in ioMRI is man-
datory to avoid confusion regarding the further surgical strat-
The Low-Field Experience at 0.2 T egy. In a subgroup of 182 patients who underwent 192
microsurgical operations in 63.4% the surgical resection
In 1995 a low-field C-shaped resistive magnet (Magnetom was continued due to the detection of residual tumor which
Open, Siemens AG, Medical Solutions, Erlangen, Germany) could be significantly reduced subsequently [29]. The analy-
with a lateral patient access of 240 was installed and mod- sis of the survival data showed a significant increase in
ified for intraoperative use. In the ‘‘Heidelberg concept’’ the progression-free and overall survival. This could also be
magnetic shielded cabin was installed adjacent to the neuro- demonstrated for low-grade glioma in a recent retrospective
surgical operating room and connected by a RF-shielded study [30].
sliding door. This setting allowed continued use of standard
instruments for microsurgical procedures. The detailed
description of the installation, the dedicated transport and
positioning system, and the integrated modified headholder The High-Field Experience at 1.5 T
have been published previously [4, 25]. In 1996 we demon-
strated that neuronavigation can be updated using ioMRI and In September 2008 we began performing intraoperative
that intraoperative re-registration was feasible with excellent high-field 1.5 T MR imaging at our institution. Until July
accuracy [26, 27]. Until 2008 688 patients were examined 2009 a total of 75 patients underwent ioMRI in 76 microsur-
intraoperatively, 541 for microsurgical operations, mainly gical procedures. The mean age was 5017 years (range,
for high-grade and low-grade gliomas, and 147 for interven- 5 78 years). 41 patients were male (mean age 4916.7;
tional procedure like biopsies, cyst-aspirations or abscess- range 5 78 years) and 34 female (mean age 5117.1;
drainages. Image quality at low-field 0.2T ioMRI was good range 10 77 years).
or acceptable in about 87.5% of the cases. However in 11.5% The digitally integrated neurosurgical suite (Brain-
only bad image quality was achieved or imaging was not SUITE1, BrainLAB AG, Feldkirchen, Germany) combin-
possible at all due to technical malfunction. In the systematic ing ioMRI, neuronavigation and a comprehensive OR data
analysis of the intraoperative images we identified four types management at our institution was workflow-optimized on a
Fig. 1 BrainSUITE1 ioMRI

Miyabi with MRI Magnetom
Espree 1.5T in the background.
MR compatible anesthesia
equipment is on the left. The
coloured floor coverings
represent the laminar air flow
field and fringe field distribution
at 5mT and 50 mT. Photograph
demonstrate the Miyabi
bridgeboard transfer system in
surgical position. In this position
the patients body is completely
outside the 5mT line and enables
a 360˚ access to the patient. The
8 channel phased array head coil
is adapted to the transfer shell
Lows and Highs: 15 Years of Development in Intraoperative Magnetic Resonance Imaging 19
Fig. 2 Transfer position:

Following a 90 rotation from the
surgery axis to the imaging axis
the Miyabi table is ready for
connection to the MR tabletop.
The integrated bridgeboard has
been released and arrested
previously. Now the shell and the
patient can easily be moved to the
MR Scanner. The entire transfer
procedure takes about 3 minutes
one room concept with an surface area of 64 m2. This setup patient the integrated and retracted bridgeboard (length 60 cm)
is functionally compartmentalized along two perpendicular has to be released and connected to the MRI-tabletop. Then the
axes in areas which are dedicated to the field of surgery, shell can easily be moved to the imaging position. The entire
anesthesia and imaging. These axes are defined by the posi- procedure takes about 3 min in average in both directions and
tion of OR table related to the MR scanner (Figs. 1, 2). In we encountered no transfer related complication. Intraopera-
surgical position the patients body is completely outside the tive imaging protocol was determined by the underlying pa-
5 mT line and provides a 360 access to the patient. The thology (mean time 21 min). In glioma the standard protocol
ceiling flange of the ceiling-mounted microscope (OPMI consists of isotropic 3D-MPRAGE, T2-TSE, T2-FLAIR and
Pentero C, Carl Zeiss Meditec AG, Jena, Germany) has diffusion-weighted imaging, if indicated all available ad-
been adjusted to attain a maximum coverage of the patient vanced imaging modalities and contrast administration can
by the working range of the microscope. Neuronavigational be added. Image quality was good or excellent in all cases.
components are the ceiling-mounted Vector Vision sky and Microsurgical operation for 36 high-grade and 7 low-grade
the treatment planning software iPlan-Net (BrainLAB AG, gliomas, 9 pituitary adenomas, 4 aneurysms, 2 medulloblasto-
Feldkirchen, Germany). The MR scanner (Magnetom Espree mas and 14 other were successfully completed. 4 patients were
1.5 T; Siemens AG, Medical Solutions, Erlangen, Germany) surgically treated for lesional or non-lesional epilepsy. The
is characterized by its compact length of 125 cm and an body preliminary result of our ongoing workflow-analysis suggests
coil diameter of 70 cm. The gradient system has a maximum a safe and efficient setup. The suite has been integrated into
field strength of 33 mT/m and a maximum slew rate of routine and runs on a daily basis.
100 T/ms. The fringe field distribution is symbolized by
coloured conductive floor coverings at the 5 mT (250
172 cm) and 50 mT lines (160131 cm). Through the
room control system all electrical devices can be centrally Conclusions
controlled. To bridge the distance between the MR isocenter
and the table column of about 480 cm we chose the Miyabi High-field ioMRI at 1.5 T significantly reduces imaging
bridgeboard transfer system (Trumpf Medical Systems, time, increases anatomical resolution and provides imaging
Saalfeld, Germany) with a weight limit of 160 kg which quality up to a standard that equals state of the art routine
was worldwide unique to that time. The patient is positioned diagnostic imaging. Meanwhile ioMRI has entered a stage,
on the MR-compatible Miyabi-shell, which is locked by where imaging objectives are far beyond the assessment of
clamps to the tabletop during surgery, the head is fixed in anatomy, pathology and extent of tumor resection. Recent
the five-point holder of the 8-channel phased array coil (Noras development of advanced MR sequences enables quantita-
MRI products GmbH, Hoechberg, Germany). To transfer the tive and semi-quantitative measurement of cerebral blood
20 T. Schmidt et al.
volume and flow, water movement and the chemical compo- suite: development, feasibility, safety, and preliminary experience.
sition of the tissue and therefore provides non-invasive in- Neurosurgery 63:412 424
14. Nimsky C, Ganslandt O, von Keller B, Fahlbusch R (2003) Pre
sight in brain and tumour physiology and biology liminary experience in glioma surgery with intraoperative high
intraoperatively which definitely will open a new chapter field MRI. Acta Neurochir Suppl 88:21 29
of ioMRI. 15. Pamir MN, Peker S, Ozek MM, Dincer A (2006) Intraoperative
MR imaging: preliminary results with 3 tesla MR system. Acta
Conflict of interest statement We declare that we have no conflict of Neurochir Suppl 98:97 100
interest. 16. Schulder M, Salas S, Brimacombe M, Fine P, Catrambone J,
Maniker AH, Carmel PW (2006) Cranial surgery with an expanded
compact intraoperative magnetic resonance imager. Technical
note. J Neurosurg 104:611 617
References 17. Sutherland GR, Kaibara T, Louw D, Hoult DI, Tomanek B,
Saunders J (1999) A mobile high field magnetic resonance system
1. Black PM, Moriarty T, Alexander E 3rd, Stieg P, Woodard EJ, for neurosurgery. J Neurosurg 91:804 813
Gleason PL, Martin CH, Kikinis R, Schwartz RB, Jolesz FA (1997) 18. Kanner AA, Vogelbaum MA, Mayberg MR, Weisenberger JP,
Development and implementation of intraoperative magnetic reso Barnett GH (2002) Intracranial navigation by using low field
nance imaging and its neurosurgical applications. Neurosurgery intraoperative magnetic resonance imaging: preliminary experi
41:831 842 ence. J Neurosurg 97:1115 1124
2. Jolesz FA, Blumenfeld SM (1994) Interventional use of magnetic 19. Schulder M (2009) Intracranial surgery with a compact, low field
resonance imaging. Magn Reson Q 10:85 96 strength magnetic resonance imager. Top Magn Reson Imaging
3. Schenck JF, Jolesz FA, Roemer PB, Cline HE, Lorensen WE, 19:179 189
Kikinis R, Silverman SG, Hardy CJ, Barber WD, Laskaris ET 20. Senft C, Seifert V, Hermann E, Franz K, Gasser T (2008) Useful
et al (1995) Superconducting open configuration MR imaging sys ness of intraoperative ultra low field magnetic resonance imaging
tem for image guided therapy. Radiology 195:805 814 in glioma surgery. Neurosurgery 63:257 266
4. Tronnier VM, Wirtz CR, Knauth M, Lenz G, Pastyr O, Bonsanto MM, 21. Nimsky C, Ganslandt O, Von Keller B, Romstock J, Fahlbusch R
Albert FK, Kuth R, Staubert A, Schlegel W, Sartor K, Kunze S (2004) Intraoperative high field strength MR imaging: implemen
(1997) Intraoperative diagnostic and interventional magnetic resonance tation and experience in 200 patients. Radiology 233:67 78
imaging in neurosurgery. Neurosurgery 40:891 900 22. Nimsky C, Ganslandt O, Hastreiter P, Wang R, Benner T, Sorensen AG,
5. Albert FK, Forsting M, Sartor K, Adams HP, Kunze S (1994) Early Fahlbusch R (2005) Preoperative and intraoperative diffusion
postoperative magnetic resonance imaging after resection of tensor imaging based fiber tracking in glioma surgery. Neurosur
malignant glioma: objective evaluation of residual tumor and its gery 56:130 137
influence on regrowth and prognosis. Neurosurgery 34:45 60, 23. Nimsky C, Ganslandt O, Hastreiter P, Wang R, Benner T, Sorensen
discussion 60 41 AG, Fahlbusch R (2007) Preoperative and intraoperative diffusion
6. Ammirati M, Vick N, Liao YL, Ciric I, Mikhael M (1987) Effect of tensor imaging based fiber tracking in glioma surgery. Neurosur
the extent of surgical resection on survival and quality of life in gery 61:178 185
patients with supratentorial glioblastomas and anaplastic astrocy 24. Nimsky C, Grummich P, Sorensen AG, Fahlbusch R, Ganslandt O
tomas. Neurosurgery 21:201 206 (2005) Visualization of the pyramidal tract in glioma surgery by
7. Ciric I, Ammirati M, Vick N, Mikhael M (1987) Supratentorial integrating diffusion tensor imaging in functional neuronavigation.
gliomas: surgical considerations and immediate postoperative Zentralbl Neurochir 66:133 141
results. Gross total resection versus partial resection. Neurosurgery 25. Wirtz CR, Tronnier VM, Albert FK, Knauth M, Bonsanto MM,
21:21 26 Staubert A, Pastyr O, Kunze S (1998) Modified headholder and
8. Stummer W, Pichlmeier U, Meinel T, Wiestler OD, Zanella F, operating table for intra operative MRI in neurosurgery. Neurol
Reulen HJ (2006) Fluorescence guided surgery with 5 aminolevu Res 20:658 661
linic acid for resection of malignant glioma: a randomised con 26. Wirtz CR, Bonsanto MM, Knauth M, Tronnier VM, Albert FK,
trolled multicentre phase III trial. Lancet Oncol 7:392 401 Staubert A, Kunze S (1997) Intraoperative magnetic resonance
9. Stummer W, Reulen HJ, Meinel T, Pichlmeier U, Schumacher W, imaging to update interactive navigation in neurosurgery: method
Tonn JC, Rohde V, Oppel F, Turowski B, Woiciechowsky C, and preliminary experience. Comput Aided Surg 2:172 179
Franz K, Pietsch T (2008) Extent of resection and survival in 27. Wirtz CR, Tronnier VM, Bonsanto MM, Knauth M, Staubert A,
glioblastoma multiforme: identification of and adjustment for Albert FK, Kunze S (1997) Image guided neurosurgery with
bias. Neurosurgery 62:564 576 intraoperative MRI: update of frameless stereotaxy and radicality
10. Wirtz CR, Albert FK, Schwaderer M, Heuer C, Staubert A, control. Stereotact Funct Neurosurg 68:39 43
Tronnier VM, Knauth M, Kunze S (2000) The benefit of neurona 28. Knauth M, Aras N, Wirtz CR, Dorfler A, Engelhorn T, Sartor K
vigation for neurosurgery analyzed by its impact on glioblastoma (1999) Surgically induced intracranial contrast enhancement:
surgery. Neurol Res 22:354 360 potential source of diagnostic error in intraoperative MR imaging.
11. Kollias SS, Bernays R, Marugg RA, Romanowski B, Yonekawa Y, AJNR Am J Neuroradiol 20:1547 1553
Valavanis A (1998) Target definition and trajectory optimization 29. Wirtz CR, Knauth M, Stamov M, Bonsanto M, Metzner R, Kunze
for interactive MR guided biopsies of brain tumors in an open S, Tronnier VM (2002) Clinical impact of intraoperative magnetic
configuration MRI system. J Magn Reson Imaging 8:143 159 resonance imaging on central nervous system neoplasia. Tech
12. Hall WA, Galicich W, Bergman T, Truwit CL (2006) 3 Tesla intrao Neurosurg 7:326 331
perative MR imaging for neurosurgery. J Neurooncol 77:297 303 30. Ahmadi R, Dictus C, Hartmann C, Zurn O, Edler L, Hartmann M,
13. Jankovski A, Francotte F, Vaz G, Fomekong E, Duprez T, Combs S, Herold Mende C, Wirtz CR, Unterberg A (2009) Long
Van Boven M, Docquier MA, Hermoye L, Cosnard G, Raftopoulos term outcome and survival of surgically treated supratentorial
C (2008) Intraoperative magnetic resonance imaging at 3 T using a low grade glioma in adult patients. Acta Neurochir (Wien) 151
dual independent operating room magnetic resonance imaging (11):1359 1365
Intraoperative Imaging in Neurosurgery: Where Will the Future
Take Us?
Ferenc A. Jolesz
Abstract Intraoperative MRI (ioMRI) dates back to the Keywords Image guided therapy Imaging Interventional
1990s and since then has been successfully applied in neu- Intraoperative MRI (ioMRI) Magnetic resonance MR
rosurgery for three primary reasons with the last one becom- Neurosurgery
ing the most significant today: (1) brain shift-corrected
navigation, (2) monitoring/controlling thermal ablations,
and (3) identifying residual tumor for resection. IoMRI, Introduction
which today is moving into other applications, including
treatment of vasculature and the spine, requires advanced
Since the early 1990s when we introduced intraoperative
3 T MRI platforms for faster and more flexible image acqui-
MRI (ioMRI) to the field of neurosurgery, it has greatly
sitions, higher image quality, and better spatial and temporal
developed and the number of users and clinical applications
resolution; functional capabilities including fMRI and DTI;
have increased [1 10]. It is now important to look back and
non-rigid registration algorithms to register pre- and intrao-
evaluate the progress of ioMRI to assess what has happened
perative images; non-MRI imaging improvements to contin-
during almost two decades of activity and to identify to the
uously monitor brain shift to identify when a new 3D MRI
best of our abilities where we are headed.
data set is needed intraoperatively; more integration of
The original reasons to develop ioMRI were the follow-
imaging and MRI-compatible navigational and robot-
ing:
assisted systems; and greater computational capabilities to
handle the processing of data. The Brigham and Women’s 1. Brain shift-corrected navigation: The navigational sys-
Hospital’s ‘‘AMIGO’’ suite is described as a setting for tems used in operating rooms rely on preoperative images
progress to continue in ioMRI by incorporating other mod- that do not reflect changes in brain anatomy due to defor-
alities including molecular imaging. A call to action is made mations and shifts during surgery. This situation has
to have other researchers and clinicians in the field of image caused inaccurate targeting and major limitations for
guided therapy to work together to integrate imaging with neuronavigation. Today, by updating information on the
therapy delivery systems (such as laser, MRgFUS, endo- brain using a 3D image database, navigation is made
scopic, and robotic surgery devices). more accurate throughout the entire surgical procedure
[11, 12].
2. Monitoring/controlling thermal ablations: The original idea
of MRI-guided interstitial laser brain surgery leveraged the
temperature sensitivity of MRI to allow for temperature
monitoring during the procedure [13 20]. Today, radio-
frequency (RF) and focused ultrasound thermal ablations
can similarly be controlled through MRI [21 26].
3. Identifying residual tumor: MRI allows the clinician to
verify completeness of tumor removal at the end of sur-
F.A. Jolesz gery and, if needed, to perform an additional tumor
B. Leonard Holman Professor of Radiology, Division of MRI and resection. The original vision for ioMRI focused on correct-
Image Guided Therapy Program, Department of Radiology, Brigham
ing for brain shift and monitoring temperature, while
and Women’s Hospital, Harvard Medical School, 75 Francis Street,
Boston, MA 02115, USA today most users apply to perform ioMRI for the third
e mail: jolesz@bwh.harvard.edu reason: tumor control [4, 27 29].
22 F.A. Jolesz
Since its introduction, ioMRI, conceptually and in real endovascular procedures. Intraoperative imaging of vascular
practice, has grown and changed significantly. anatomy, including feeding arteries and the draining veins
of an intracranial arteriovenous malformation (AVM),
allows the neurosurgeon to intraoperatively assess vessels
and surrounding areas and make decisions about the need
Present Benefits and Capabilities to completely and safely resect or obliterate a vascular
malformation.
The original full access open MRI, like the SIGNA SP Various diseases of the spine: Spinal surgeries with IMR
(General Electric Healthcare Technologies, Waukesha, will include: lumbar discectomies, anterior cervical discec-
WI), was replaced with open limited access low and midfield tomies with fusion, cervical vertebrectomies, foramino-
magnets and high (1.5 T) and ultra high field (3 T) closed tomies, and laminectomies. Through ioMRI, the accuracy
bore MRI scanners [6, 9, 10, 30 32]. Although, the SIGNA of localization and adequacy of decompression can be
SP represented the best configuration, it is at midfield and, assessed. Future applications may also include: resection of
therefore, not optimal for neurosurgery. Today most neuro- spine and spinal cord, tumors, and spinal endoscopy. Tem-
surgeons want higher field strength and more advanced perature sensitive imaging for laser discectomy and verteb-
image acquisition technology that is only available in high roplasty can be performed with ioMRI guidance to prevent
field advanced systems to obtain higher image quality and the thermal damage of the nerve roots and spinal cord.
better spatial and temporal resolution.
Advanced 3 T imaging platforms achieve faster and more
flexible image acquisitions that improve localization, target- ioMRI Approaches
ing, monitoring, and therapy control. Advance imaging
methods functional MRI (fMRI) and diffusion tensor imag-
ing (DTI) have been introduced into neurosurgery and The ‘‘tumor control’’ method of ioMRI is a practical and well
integrated into intraoperative imaging [33 35]. accepted use of the technology, though it is best applied, not
As far as clinical applications are concerned, ioMRI has as a single imaging session at the completion of surgery
been successfully developed and implemented for multiple when the clinician cannot take full advantage of ioMRI,
procedures, including: but throughout surgery to provide corrected, brain shift-
compensated guidance (with serial imaging and navigation).
• Biopsies and placement of electrodes The goal of more complete and effective tumor resection can
• Craniotomies for image-guided resection of benign and only be accomplished if we have better techniques to detect
malignant intracranial tumors the full extent of infiltrative brain tumors. ioMRI is part of
• Intra-cranial cyst drainages and this solution along with other multimodality-based molecu-
• Thermal ablations for malignant and benign tumors. lar imaging (optical, nuclear, or other). Besides tumor mar-
gin detection, we need more detailed information about the
functional and structural anatomy of the normal brain tissue
that surrounds the tumor (for MRI such detail comes through
Future ioMRI Applications multiparametric imaging: T1, T2, fMRI, DTI perfusion
imaging).
In the future, this application pool will expand to:
Bening tumors (skull base): MRI-enhanced neuroendo-
scopy can be faster and safer than conventional neuroendo-
scopy and minimize the possibility of complications. Managing the Brain Shift Challenge
A strong need exists to develop MRI-compatible flexible
endoscopes with position tracking coils. The combination Navigational systems fully integrated with ioMRI are also
of the two techniques may provide the best surgical guid- part of improved image guidance [8, 11, 36 39]. For exam-
ance. The most promising application is endoscopic trans- ple, it is necessary to monitor brain shift with a non-MRI
phenoidal surgery for midline skull base and parasellar technique for continuous monitoring that helps the surgeon/
lesions and transventricular endoscopic removal for extra- neuroradiologist to decide when a new MRI 3D dataset
ventricular tumors. should be obtained. Development of an integrated system
Neurovascular abnormalities and stroke: ioMRI provides to improve visualization, navigation, and monitoring is a
real-time perfusion and diffusion imaging to intraopera- necessary step in this direction.
tively diagnose acutely developed vascular occlusion and to Today we know it is possible to create an augmented
monitor the condition of the brain during surgeries and/or reality visualization of the intraoperative configuration of
Intraoperative Imaging in Neurosurgery: Where Will the Future Take Us? 23
the patient’s brain merged with high resolution preoperative (MRgFUS) as well as the targeted ablative treatment of brain
imaging data, including DTI and fMRI to better localize tumors via drug delivery through an open blood-brain-barrier
the tumor and critical healthy tissues. Brain shift-corrected (BBB) [23 25, 52 57]. FUS-induced neuromodulation is a
imaging requires the use of non-rigid registration algorithms complement to functional neurosurgical applications for coa-
to compensate for displacements [40 42]. Massive compu- gulative lesions [57 59].
tational needs must support these capabilities that entail Greater progress and expansion for ioMRI requires the
online usage in reasonable time frames of less than 5 min. development of new, more advanced imaging methods,
One of the main reasons for this computational load is that, navigational techniques, surgical instruments and tools; the
to properly account for ongoing brain shifts, fMRI and DTI more efficient use of computing technologies; and the inte-
must be accurately co-registered and updated. gration of diagnostic and therapy devices with navigational
Methods of monitoring brain shift now include: tools (computer-assisted surgery). A need exists for a multi-
Laser Range Scanner (LRS): Devices, mostly used in focused, multidisciplinary effort involving researchers and
robotics applications, that emit and receive a laser beam clinicians to explore and refine ioMRI to make it as cost-
and by measuring difference of phase, time of flight or effective and as widely accessible to a multidisciplinary pool
frequency shift depth is measured. of users.
Stereo camera imaging: Imaging from a type of camera
with two or more lenses that can capture three-dimensional
images. AMIGO: Merging ioMRI and Other Modalities
Transcranial US (ITUM): A type of device, at the proto-
type stage, that uses the shear mode of transcranial ultra- At the Brigham and Women’s Hospital, we are taking a lead
sound transmission to intraoperatively monitor brain shift. in this effort through the development and execution of a
MRI can then be obtained with the device spatially multimodality approach to intraoperative imaging by inte-
registered to the MRI reference coordinates [43]. grating the components of a rich multimodal and translational
clinical research environment to carry out open and mini-
mally invasive surgeries and percutaneous interventions.
Multimodality IGT provides comprehensive information
Further Development derived from different physical and biological properties.
Combined multimodalities provide anatomical, metabolic,
ioMRI allows for the introduction of new surgical and functional information a previously stated goal for
approaches and/or techniques in vascular neurosurgery, ioMRI. Images can be complemented with newly developed
spine surgery, and skull base surgery (using image guided multiple molecular probes (nuclear, optical, mass spectrom-
and registered neuroendoscopy). Further moving ioMRI eter, etc.). The use of multiple molecular probes improves
forward in neurosurgery is the already-underway integra- the sensitivity and specificity of cancer-relevant applications
tion of ioMRI with image guided robots that provide, over single mode imaging. To achieve this we must expand
through remote manipulation, an alternative solution to limit- our imaging platform for image guided neurosurgery to
ed access. Already, for many types of neurosurgical pro- include all modalities, especially those applicable to molec-
cedures, companies have developed MRI-compatible ular imaging (PETCT, MRI, optical imaging).
robotic systems and manipulators [44 51]. Today most The validation of imaging tools and molecular imaging
systems still lack the clinical validation that is necessary agents in a multimodal surgical setting is possible because,
for the surgical robot to execute not only biopsies but also during surgeries, multiple tissue samples can be obtained for
complex surgical manipulations. pathology that can validate imaging findings at the same
ioMRI today has also moved beyond MRI-guided and location. Navigational and registration methods for multi-
controlled thermal ablations for interstitial laser therapy [13, modal registration can be used to compare multiple probes
14, 16, 17, 19] to MRI-guided focused ultrasound surgery that are measuring exactly the same tissue region.
Fig. 1 Advanced Multimodality Image Guided OR (AMIGO). The middle room is an operating room with surgical microscope, angiography,
ultarsound and optical imaging. The room on the left has a PETCT and on the right a 3T MRI which is seiling mounted and can move into the
operating room to image the patient on the operating room table (IMRIS, Winnipeg, Canada)
24 F.A. Jolesz
The BWH’s Advanced Multimodality Image Guided 11. Jolesz FA, Kikinis R, Talos IF (2001) Neuronavigation in inter
Operating (AMIGO) Suite (see Fig. 1), opening in 2010, ventional MR imaging. Frameless stereotaxy. Neuroimaging Clin
N Am 11(4):685 689
is uniquely designed to support, alongside routine clinical 12. Nabavi A, Black PM, Gering DT et al (2001) Serial intraoperative
procedures, the next phase of clinical and research activity in magnetic resonance imaging of brain shift. Neurosurgery 48
ioMRI that will incorporate multimodal imaging. The logis- (4):787 797, discussion 797 798
tics of the suite are such that multidisciplinary teams will be 13. Anzai Y, Lufkin R, DeSalles A, Hamilton DR, Farahani K,
Black KL (1995) Preliminary experience with MR guided thermal
in close cooperation with one another during procedures that ablation of brain tumors. AJNR Am J Neuroradiol 16(1):39 48,
might involve the use of an array of on-site (i.e., in the suite discussion 49 52
itself) modalities from 3D ultrasound, fluoroscopy, MRI, and 14. Ascher PW, Justich E, Schröttner O (1991) Interstitial thermo
PET/CT to carry out a variety of procedures. therapy of central brain tumors with the Nd:YAG laser under
real time monitoring by MRI. J Clin Laser Med Surg 9(1):79 83
The AMIGO and suites like it are at the forefront by 15. Bettag M, Ulrich F, Schober R et al (1991) Stereotactic laser
providing settings for practice, refinement, improvement, therapy in cerebral gliomas. Acta Neurochir Suppl (Wien)
validation, and innovation. Indeed, the future of intraopera- 52:81 83
tive imaging in neurosurgery depends on such settings and 16. Fan M, Ascher PW, Schröttner O, Ebner F, Germann RH, Kleinert R
(1992) Interstitial 1.06 Nd:YAG laser thermotherapy for brain
teams and their increased application of advanced technolo- tumors under real time monitoring of MRI: experimental study
gies and multimodality imaging. Integration of imaging with and phase I clinical trial. J Clin Laser Med Surg 10(5):355 356
therapy delivery systems (laser, MRgFUS, endoscopic and 17. Kahn T, Bettag M, Ulrich F et al (1994) MRI guided laser induced
robotic surgery devices) will characterize the next phase of interstitial thermotherapy of cerebral neoplasms. J Comput Assist
Tomogr 18:519 532
this emerging field. 18. Kettenbach J, Silverman SG, Hata N et al (1998) Monitoring and
visualization techniques for MR guided laser ablations in an open
Conflict of interest statement Dr. Jolesz has received cosultant fees MR system. J Magn Reson Imaging 8(4):933 943
from General Electric, who pioneered this system. 19. McDannold NJ, Jolesz FA (2000) Magnetic resonance image
guided thermal ablations. Top Magn Reson Imaging 11(3):
191 202
20. Stollberger R, Ascher PW, Huber D, Renhart W, Radner H, Ebner F
(1998) Temperature monitoring of interstitial thermal tissue
References coagulation using MR phase images. J Magn Reson Imaging
8(1):188 196
21. Cline HE, Hynynen K, Watkins RD et al (1995) Focused US
1. Black PM, Alexander E 3rd, Martin C et al (1999) Craniotomy for system for MR imaging guided tumor ablation. Radiology 194
tumor treatment in an intraoperative magnetic resonance imaging (3):731 737
unit. Neurosurgery 45(3):423 431 22. Hynynen K, Vykhodtseva NI, Chung AH, Sorrentino V, Colucci V,
2. Dimaio SP, Archip N, Hata N et al (2006) Image guided neuro Jolesz FA (1997) Thermal effects of focused ultrasound on the brain:
surgery at Brigham and Women’s Hospital. IEEE Eng Med Biol determination with MR imaging. Radiology 204(1):247 253
Mag 25(5):67 73 23. Jolesz FA, Hynynen K (2002) Magnetic resonance image guided
3. Hall WA, Martin AJ, Liu H et al (1998) High field strength inter focused ultrasound surgery. Cancer J 8(suppl 1):S100 S112
ventional magnetic resonance imaging for pediatric neurosurgery. 24. Jolesz FA, Hynynen K, McDannold N, Tempany C (2005) MR imaging
Pediatr Neurosurg 29(5):253 259 controlled focused ultrasound ablation: a noninvasive image
4. Jolesz FA, Talos IF, Schwartz RB et al (2002) Intraoperative guided surgery. Magn Reson Imaging Clin N Am 13(3):545 60
magnetic resonance imaging and magnetic resonance imaging 25. Jolesz FA, McDannold N (2008) Current status and future potential
guided therapy for brain tumors. Neuroimaging Clin N Am 12 of MRI guided focused ultrasound surgery. J Magn Reson Imaging
(4):665 683 27(2):391 399
5. Lewin JS, Metzger AK (2001) Intraoperative MR systems. Low 26. McDannold N, Hynynen K, Wolf D, Wolf G, Jolesz F (1998) MRI
field approaches. Neuroimaging Clin N Am 11(4):611 628 evaluation of thermal ablation of tumors with focused ultrasound.
6. Schenck JF, Jolesz FA, Roemer PB, Cline HE, Lorensen WE, J Magn Reson Imaging 8(1):91 100
Kikinis R et al (1995) Superconducting open configuration 27. Claus EB, Horlacher A, Hsu L (2005) Survival rates in patients
MR imaging system for image guided therapy. Radiology 195 with low grade glioma after intraoperative magnetic resonance
(3):805 814 image guidance. Cancer 103(6):1227 1233
7. Schwartz RB, Hsu L, Wong TZ et al (1999) Intraoperative MR 28. Mittal S, Black PM (2006) Intraoperative magnetic resonance
imaging guidance for intracranial neurosurgery: experience with imaging in neurosurgery: the Brigham concept. Acta Neurochir
the first 200 cases. Radiology 211(2):477 488 Suppl 98:77 86
8. Schulder M, Liang D, Carmel PW (2001) Cranial surgery naviga 29. Nimsky C, Fujita A, Ganslandt O, Von Keller B, Fahlbusch R
tion aided by a compact intraoperative magnetic resonance imager. (2004) Volumetric assessment of glioma removal by intraoperative
J Neurosurg 94(6):936 945 high field magnetic resonance imaging. Neurosurgery 55(2):
9. Steinmeier R, Fahlbusch R, Ganslandt O et al (1998) Intraoperative 358 370, discussion 370 371
magnetic resonance imaging with the magnetom open scanner: 30. Bradley WG (2002) Achieving gross total resection of brain
concepts, neurosurgical indications, and procedures: a preliminary tumors: intraoperative MR imaging can make a big difference.
report. Neurosurgery 43(4):739 747, discussion 747 748 AJNR Am J Neuroradiol 23(3):348 349
10. Sutherland GR, Kaibara T, Louw D, Hoult DI, Tomanek B, 31. Nimsky C, Ganslandt O, Von Keller B, Romstöck J, Fahlbusch R
Saunders J (1999) A mobile high field magnetic resonance system (2004) Intraoperative high field strength MR imaging: implemen
for neurosurgery. J Neurosurg 91(5):804 813 tation and experience in 200 patients. Radiology 233(1):67 78
Intraoperative Imaging in Neurosurgery: Where Will the Future Take Us? 25
32. Truwit CL, Hall WA (2006) Intraoperative magnetic resonance 46. DiMaio SP, Pieper S, Chinzei K et al (2007) Robot assisted needle
imaging guided neurosurgery at 3 T. Neurosurgery 58(4 suppl 2): placement in open MRI: system architecture, integration and vali
ONS 338 345, discussion ONS 345 346 dation. Comput Aided Surg 12(1):15 24
33. Hall WA, Liu H, Truwit CL (2005) Functional magnetic resonance 47. Elhawary H, Tse ZT, Hamed A, Rea M, Davies BL, Lamperth MU
imaging guided resection of low grade gliomas. Surg Neurol (2008) The case for MR compatible robotics: a review of the state
64(1):20 27, discussion of the art. Int J Med Robot 4(2):105 113
34. Golby AJ, McConnell KA (2004) Functional brain mapping 48. Elhawary H, Zivanovic A, Davies B, Lampérth M (2006) A review
options for minimally invasive surgery. In: Black PM, Proctor M of magnetic resonance imaging compatible manipulators in sur
(eds) Minimally invasive neurosurgery. Humana Press, Totawa, NJ, gery. Proc Inst Mech Eng 220(3):413 424
pp 87 106 49. Masamune K, Kobayashi E (1995) Development of an MRI
35. Nimsky C, Ganslandt O, Kober H et al (1999) Integration of compatible needle insertion manipulator for stereotactic neurosur
functional magnetic resonance imaging supported by magnetoen gery. J Image Guid Surg 1(4):242 248
cephalography in functional neuronavigation. Neurosurgery 44 50. Sutherland GR, Latour I, Greer AD (2008) Integrating an image
(6):1249 1255, discussion 1255 1256 guided robot with intraoperative MRI: a review of the design and
36. Gering DT, Nabavi A, Kikinis R et al (2001) An integrated visuali construction of neuroArm. IEEE Eng Med Biol Mag 27(3):59 65
zation system for surgical planning and guidance using image 51. Sutherland GR, Latour I, Greer AD, Fielding T, Feil G, Newhook P
fusion and an open MR. J Magn Reson Imaging 13(6):967 975 (2008) An image guided magnetic resonance compatible surgical
37. Nabavi A, Gering DT, Kacher DF et al (2003) Surgical navigation robot. Neurosurgery 62(2):286 292, discussion 292 293
in the open MRI. Acta Neurochir Suppl 85:121 125 52. Hynynen K, McDannold N, Vykhodtseva N, Jolesz FA (2001)
38. Nimsky C, Ganslandt O, Cerny S, Hastreiter P, Greiner G, Noninvasive MR imaging guided focal opening of the blood
Fahlbusch R (2000) Quantification of, visualization of, and com brain barrier in rabbits. Radiology 220(3):640 646
pensation for brain shift using intraoperative magnetic resonance 53. Kinoshita M, McDannold N, Jolesz FA, Hynynen K (2006)
imaging. Neurosurgery 47(5):1070 1079, discussion 1079 1080 Targeted delivery of antibodies through the blood brain barrier
39. Nimsky C, Ganslandt O, Hastreiter P, Fahlbusch R (2001) Intrao by MRI guided focused ultrasound. Biochem Biophys Res Com
perative compensation for brain shift. Surg Neurol 56(6):357 364, mun 340(4):1085 1090
discussion 364 365 54. Kinoshita M, McDannold N, Jolesz FA, Hynynen K (2006) Non
40. Clatz O, Delingette H, Talos IF et al (2005) Hybrid formulation of invasive localized delivery of Herceptin to the mouse brain by
the model based non rigid registration problem to improve accuracy MRI guided focused ultrasound induced blood brain barrier dis
and robustness. Med Image Comput Comput Assist Interv ruption. Proc Natl Acad Sci U S A 103(31):11719 11723
8(Pt 2):295 302 55. Treat LH, McDannold N, Vykhodtseva N, Zhang Y, Tam K,
41. Ferrant M, Nabavi A, Macq B et al (2002) Serial registration Hynynen K (2007) Targeted delivery of doxorubicin to the rat
of intraoperative MR images of the brain. Med Image Anal brain at therapeutic levels using MRI guided focused ultrasound.
6(4):337 359 Int J Cancer 121(4):901 907
42. Warfield SK, Haker SJ, Talos IF et al (2005) Capturing intraopera 56. Vykhodtseva N, McDannold N, Hynynen K (2006) Induction of
tive deformations: research experience at Brigham and Women’s apoptosis in vivo in the rabbit brain with focused ultrasound and
Hospital. Med Image Anal 9(2):145 162 optison. Ultrasound Med Biol 32(12):1923 1929
43. White PJ, Whalen S, Tang SC, Clement G, Jolesz F, Golby AJ 57. Yoo SS, Lee JH, Zhang Y et al (2008) FUS mediated reversible
(2009) An intraoperative brain shift monitor using shear mode modulation of region specific brain function. Proc MRgFUS
transcranial ultrasound: preliminary results. J Ultrasound Med 2008:10
28(2):191 203 58. Colucci V, Strichartz G, Jolesz F, Vykhodtseva N, Hynynen K
44. Chinzei K, Miller K (2001) Towards MRI guided surgical manip (2009) Focused ultrasound effects on nerve action potential
ulator. Med Sci Monit 7(1):153 163 in vitro. Ultrasound Med Biol 35(10):1737 1747
45. Chinzei K, Warfield S, Hata N, Tempany C, Jolesz F, Kikinis R 59. Currier DP, Greathouse D, Swift T (1978) Sensory nerve con
(2003) Planning, simulation and assistance with intraoperative duction: effect of ultrasound. Arch Phys Med Rehabil 59(4):
MRI. Minim Invasive Ther Allied Technol 12(1):59 64 181 185
Intraoperative MRI- Ultra Low Field Systems
Development and Design of Low Field Compact
Intraoperative MRI for Standard Operating Room
Moshe Hadani
Abstract Objectives: To present the development of a com- Introduction

pact low field intraoperative MR image guidance system and
its application in brain surgery. Intraoperative magnetic resonance imaging (ioMRI) was
Methods: The PoleStar ioMRI system (Odin Medical introduced in 1997 to complement standard neuronavigation
Technologies, Israel and Medtronic, Inc. USA) was devel- with updated, high resolution MR images during surgery [1]
oped for use in a standard operating room. Its primary The first system was a 0.5 T MRI scanner (GE Signa SP)
physical fixed parameters are magnetic field of 0.15 T and which weighted 6,000 kg and was installed in a special
field of view of 2016cm. The magnet is mounted on a room, outside of the operating rooms complex. Our site,
transportable gantry and can be positioned under the surgical among a few others, had the opportunity to operate the
table when not in use for scanning. Additional functionality system. The surgical environment was different from the
includes integrated navigation, and system operation by the standard. Working outside of the operating room complex
surgeons. was problematic and the cost of the magnet and the building
Results: The PoleStar system integrates into existing was high. At the same time a group of MRI physicists and
operating rooms requiring only slight modification of the engineers in Israel developed a small, low field MRI scanner
surgical environment. Standard instruments can be used. of 0.12 T, which weighted only 500 kg. The challenge was to
The system’s imaging allows it to be used for the following bring this scanner to function in a standard neurosurgery
indications: pituitary tumors, low grade gliomas (including operating room.
awake surgery), high grade gliomas, intraventricular tumors, The goals for the clinical design of system were:
accurate navigation to small lesions such as cavernous
• Provide optimal MR imaging for navigation and tumor
angiomas or metastases, drainage of cysts and brain abscesses.
resection control.
The image quality, which is comparable to post operative
• Integrated navigation capabilities
diagnostic high field imaging, enables high quality resection
• Integration into existing operating rooms, allowing the
control.
use of standard surgical tools and equipment.
More than 6,000 brain surgeries were done with the
• Minimize interference with routine surgical procedures.
system in 50 centers in the US and Europe.
• No requirement for structural changes to the building and
Conclusion: The low field intraoperative MRI system is a
low cost of purchase and installation.
valuable tool in the modern operating room.
• Provide full control by the surgeon including MRI
scanning.
Keywords Compact ioMRI Image-guided surgery ioMRI
Intraoperative magnetic resonance imaging Low field MRI
Resection control Surgical neuronavigation
Materials and Methods
The following were the fixed physical parameters of the

PoleStar scanner: Low field magnetic strength (0.15 T in
the PoleStar N20, which followed the original PoleStar
M. Hadani
N-10, Odin Medical Technologies Israel and Medtronic
Department of Neurosurgery, Sheba Medical Center, Sackler School of
Medicine, Tel Aviv University, Tel Hashomer 52621, Israel inc USA) The scanner consists of two vertical, parallel,
e mail: moshe.hadani@sheba.health.gov.il disk-shaped permanent magnets located 27 cm from each
30 M. Hadani
Fig. 1 The compare function shows consecutive T1 weighted axial sections of a patient with a grade II astrocytoma, illustrating occurring brain
shift of 1.5 cm and residual tumor, that was located by intraoperative navigation and resected completely
other. This is the space for the positioning of the head. The with different sequences, such as T1 and T2 contrasts
field-of-view is 2016 cm. The magnets are affixed to a (Fig. 1).
U-shaped arm. The arm itself is mounted on a transportable The scanner is integrated with a navigation station.
gantry positioned below the operating table. For scanning Acquired images are immediately available for navigation.
the magnets are raised, by the system computer under tight Patient registration is not required. Navigation on diagnostic
manual control, so that the area to be imaged is well centered images is also supported.
between the magnet poles. After scanning, the magnets are
retracted under the OR table providing good access to the
head during surgery.
A special MRI compatible head holder was designed to Results
accommodate the limited space for the head between the
magnet poles. Due to the low field and the compact size the scanner was
The following MRI sequences were developed: T1-weighted, installed in a standard operating room (Fig. 2). The only
T2-weighted, FLAIR and e-steady sequences. Image acquisition modification of the room was the installation of copper
takes from 8 s to 13 min, depending on the selected MR sequence mesh in the walls to eliminate RF noise. Light, power and
and slice thickness. Typically, 7-min T1-weighted images and anesthesiology gas inlets are filtered to avoid interference
4 mm slice thickness are used as preoperative baseline with image acquisition.
imaging. When not in use, the scanner is placed in a special cabinet
Image series from different scans can be displayed to allow the room to be used for any surgery. The low
side by side, facilitating comparison between different magnetic field of the system is largely confined to the area
stages of surgery as well as comparison of images acquired between the magnets and decays to 5 G 1.7 m from the
Development and Design of Low Field Compact Intraoperative MRI for Standard Operating Room 31
All equipment and tools are the standard. This includes

conventional operating tables, microscopes, chairs, ultrasonic
aspirator, standard drill, bipolar stimulator, EEG recorder.
MRI scans are controlled by the surgeon or an assistant at
the navigation workstation (Treon, Medtronic, Inc.)
Navigation is done on the last MRI set. No registration is
needed, the magnet itself, rather than the patient’s head, is
automatically registered, and the accuracy is high [2].
More than 600 adult and pediatric cases were operated
with the PoleStar ioMRI system in our center in Tel Aviv.
Currently about 6,000 operations were done with the system
worldwide.
Clinical applications included craniotomies for intraxial
and extraaxial brain tumors, metastastic tumors, pituitary
tumors, intraventricular lesions, cavernous angiomas, shunt
placements, epilepsy surgery, awake surgeries, stereo tactic
biopsies and cyst aspirations.
Discussion
Following the 10 years experience with ioMRI, there are

currently 90 systems installed worldwide: about 40 high
magnetic field systems (1.5 T and 3 T), 50 low field systems
(0.15 T) and a few 0.3 T systems.
8,000 9,000 surgeries were performed with ioMRI.
The utility of ioMRI is mainly resection control, correc-
tion of brain shift, complication avoidance, and navigation to
small lesions.
The design goals of the PoleStar compact system were
met [3 5]. Its intraoperative images serve as the data for
navigation and for resection control. The image quality of
the low field system is inferior to that of the high field
systems; however, it was found optimal for updating the
navigation data and for the ability to navigate with high
accuracy to small lesions. The PoleStar imaging enables
high quality resection control, which is comparable to post-
operative diagnostic high field imaging (Fig. 3).
All patients’ positions are possible and manipulation of
the operative table enables the surgeon to work in the most
comfortable position.
Fig. 2 (a) The PoleStar system in a standard operating room shielded For the anesthesiologist, [6] using the PoleStar ioMRI
for RF. On the right the ‘‘cabinet’’ for the magnet when not in use. The
system allows preservation of working conditions that are
scanner is integrated with the Treon navigation system. (b) Operation
room setup. Only anesthesia equipment is MRI compatible similar to a regular operating room. The patient is accessible
to the anesthesiologist at all times during the procedure, and
conventional syringe pumps and warming devices can be
magnet’s isocenter. When the scanner is placed below the used. It is also possible to perform electrophysiological
level of the OR table, the magnetic field strength in the monitoring in surgery for epilepsy, and to carry out surgery
operative field is below 50 G, enabling the use of standard under intravenous sedation and intra-operative mapping in
surgical instruments. the awake state.
32 M. Hadani
Fig. 3 Intraoperative extended resection of a pituitary macro adenoma. Compare mode
In exchange for the ‘‘lower quality’’ imaging the follow- Conclusions

ing advantages were achieved in the development of the
PoleStar system: The development of a low field compact intraoperative MRI
Easy to use in a standard operating room with no need to system was custom made to the clinical need.
change the normal flow of the operative procedure. This low field ioMRI system functions in a normal
Safety, when using the system in a standard operating operating room modified only for radiofrequency interfer-
room environment. ence. The operative environment is normal and standard
Keeping the duration of the operation within the usual instruments are used. Use of the system does not change
timeframe with only minor addition of time, and last but not the working habits in the OR.
least The scanning and navigation capabilities of the system
The compact system is by far less expansive due to the eliminate the inaccuracies that may result from brain shift.
lower cost of the components of the system and the low cost Image quality enables high quality resection control, which
of installation in the operation room without the need for is comparable to post operative diagnostic high field imaging.
special building. The main indications for the use of ioMRI are: pituitary
ioMRI should be part of the modalities used in brain tumors, low grade gliomas (awake), high grade gliomas,
surgery. It is useful in a subset of operations such as: intraventricular tumors, accurate navigation to small lesions
pituitary tumor surgery, high and low grade gliomas, for such as cavernous angiomas or metastases, drainage of cysts
resection control and complication avoidance, Navigation and brain abscesses.
and resection of small lesions 1 cm or less such as cavernous
angiomas or small metastatic tumors, drainage of cysts and Conflict of interest statement We declare that we have no conflict of
interest.
abscesses, navigation during facedown procedures.
Development and Design of Low Field Compact Intraoperative MRI for Standard Operating Room 33
References guided system for conventional neurosurgical operating rooms.

Neurosurgery 48(4):799 809
1. Black PM, Moriarty T, Alexander E 3rd, Steig P, Woodard EJ, 4. Hadani M, Schulder M, Bernays RL (2002) Compact 0.12 tesla
Gleason L, Martin CH, Kikinis R, Schwartz RB, Jolesz F (1997) intraoperative magnetic resonance image guidance system in the
Development and implementation of intraoperative magnetic standard operating room. Tech Neurosurg 7(4)
resonance imaging and its neurosurgical applications. Neurosur 5. Schulder M, Salas S, Brimacombe M, Fine P, Catrambone J,
gery 41(4):831 845 Maniker A, Carmel P (2006) Cranial surgery with an expanded
2. Salas S, Brimacombe M, Schulder M (2007) Stereotacti accuracy compact intraoperative magnetic resonance imager. J Neurosurg
of a compact intraoperative MRI system. Stereotact Funct Neuro 104:611 617
surg 85:69 74 6. Berkenstadt H, Perel A, Ram Z, Feldman Z, Nahtomi Shick O,
3. Hadani M, Spiegelman R, Feldman Z, Berkenstadt H, Ram Z Hadani M (2001) Anesthesia for magnetic resonance guided neu
(2001) Novel, compact, intraoperative magnetic resonance imaging rosurgery. Initial experience with a new open magnetic resonance
imaging system. J Neurosurg Anesthesiol 13:158 162
Low Field Intraoperative MRI in Glioma Surgery
Volker Seifert, Thomas Gasser, and Christian Senft
Abstract The extent of resection marks one prognostic fac- imaging modalities, recent studies could show that a radio-
tor for patients with malignant gliomas. Among the methods logically complete resection represents an independent
used for the intraoperative control of the extent of resec- prognostic factor for patients with malignant gliomas [2 6].
tion, intraoperative magnetic resonance imaging (ioMRI) In contrast to e.g. abdominal oncological surgery, brain
has become a very attractive method. It was introduced in tumor resections cannot be performed with a safety margin.
the in the final decade of the last century. The first available Due to the infiltrative nature of gliomas and the fact that their
system was a low magnetic field strength unit employing margins are frequently difficult to be visualized intraopera-
0.5 Tesla (T). While currently high-field systems (1.5 T and tively, postoperative imaging studies reveal unintentionally
above) are being developed, different low-field ioMRI sys- remaining tumor tissue quite frequently [3, 7]. To improve
tems (0.5 T and below) have been used for brain tumor the extent of resection, and thus the prognosis for glioma
resection in far more centers than high-field ioMRI, patients, neurosurgeons have sought to develop different
corresponding to a greater number of publications. Undoubt- methods of intraoperative imaging to visualize tumor tissue.
edly, high-field ioMRI systems offer superior image quality Compared to intraoperative ultrasound or intraoperative
and faster acquisition times. Yet, low-field ioMRI has influ- computed tomography, MRI offers superior image quality
enced intraoperative decision-making and improved brain without radiation exposure to the patient and/or staff. Mag-
tumor resection. With this article, we review the use of netic resonance imaging has become the gold standard for
low-field ioMRI in glioma surgery. the diagnostic and follow-up imaging of gliomas. Intrao-
perative MRI seemed as the method of choice in the resec-
Keywords Glioma surgery Intraoperative MRI Review tion control of gliomas, and it was developed for mainly, but
not exclusively, this reason [8 10].
Surgical Treatment of Gliomas

Development of Low-Field ioMRI-Systems
Rickman Godlee performed the first successful resection
Peter Black and Ferenc Jolesz from the Brigham and
of a glioma in 1884 [1]. Since then, numerous reports
Women’s Hospital in Boston pioneered intraoperative
have stated beneficial effects of extensive tumor resections
magnetic resonance imaging in the 1990s. In collaboration
despite a nonetheless deleterious course of the disease. With
with General Electric Co. they developed the first ioMRI
the introduction of microneurosurgical techniques and refined
scanner to be used in a neurosurgical operating room. The
Signa SP employed a magnetic field strength of 0.5 T and was
nicknamed ‘‘double-donut’’ system due to the shape and
V. Seifert (*) and C. Senft appearance of the system it consisted of two magnets with
Department of Neurosurgery, Johann Wolfgang Goethe University, a vertical gap in which the patient’s head was placed during
Schleusenweg 2 16, 60528 Frankfurt, Germany surgery. It required major modifications of OR-infrastructure
e mail: v.seifert@em.uni frankfurt.de
to meet the demands for low radiofrequency interference as
T. Gasser
Department of Neurosurgery, Johann Wolfgang Goethe University,
well as adaptation regarding surgical tools, instruments,
Schleusenweg 2 16, 60528 Frankfurt, Germany microscopes, etc., which all needed to be MR-compatible
University of Essen, Essen, Germany [11, 12]. While the provided images were consistently of
36 V. Seifert et al.
good quality, one major concern with this system was the into the scanner’s field of view [13, 14]. However, image
constraint working area of the surgeon, who stood in between acquisition necessitated transferring the patient from the
the magnets (Fig. 1). Also, some restrictions applied regarding operating site to the scanner, which is time-consuming.
patient positioning. While the AIRIS II MRI scanner (Hitachi Corp.), employing
Simultaneosly, Siemens Corp. developed a C-shaped a magnetic field strength of 0.3 T, was developed primarily as a
resistive MRI scanner with a static magnetic field with conventional diagnostic tool, it was also employed as an ioMRI
strength of 0.2 T (Magnetom Open). In contrast to the scanner [15, 16]. It is an open MRI unit with two horizontally
Signa SP which marked the center of the operating theater, oriented magnets with a vertical opening of 17 in. Surgeries
the Magnetom Open was set up at the one end of the opera- can be performed either in the adjacent operating room, or with
ting room, separated from the surgical area by radiofre- the patient positioned on the scanner’s table.
quency shielding, thus allowing for the use of standard The PoleStar by Odin/Medtronic Inc. is one of the most
instrumentarium during surgery. The patient had to be trans- frequently used low-field systems today with a static mag-
ferred to the magnet, which featured a 240 opening, allow- netic field strength of 0.12 (Model N-10) or 0.15 T (Model
ing for safe placement of the anaesthetized patient’s head N-20), respectively [17, 18]. It was developed to overcome
the necessity of using special, MRI-compatible instruments
and surgical devices, demanded for by other ioMRI systems,
yet avoiding cumbersome patient transfer to the scanner.
The PoleStar was designed as a mobile MRI unit with two
vertical magnets spaced 25 or 27 cm apart, respectively.
During surgery, conventional instruments can be used
while the magnet is parked underneath the operating table
(Fig. 2). The scanner is moved upwards for intraoperative
image acquisition [19, 20].
Image Quality in Low Field MRI
Although there is a span of different magnet designs, field

Fig. 1 Example of a brain tumor surgery perfomed with the Signa
SP the surgeon stands in between the magnets’ gap using non strengths, and investigational concepts, there are many simi-
ferromagnetic instruments larities in the information obtained in order to guide and
Fig. 2 Example of a brain tumor

surgery performed with the
PoleStar N 20. (a) Fixation of the
head and application of the
radiofrequency coil (b) Transfer
of the magnet to its position
underneath the operating table
(c) The magnet is moved upwards
for image acquisition (d) The
magnet is lowered and tumor
resection is performed with
standard instruments in a
microsurgical fashion
Low Field Intraoperative MRI in Glioma Surgery 37
monitor neurosurgical procedures. All of the above men- (fluid attenuated) inversion recovery (FLAIR) sequences
tioned systems have been successfully used to monitor were found to be superior in these cases [12, 30, 32].
craniotomy and extent of resection in patients with gliomas
[15, 17, 19, 21 27].
However, each has its own set of advantages and disad-
vantages, and as ioMRI-technology still evolves, no single Indications for ioMRI
system has gained universal use. Image quality in MRI is
directly linked to the field strength of the magnet and to the Implementing neuronavigation was a major step towards an
homogeneity and stability of the static and gradient magnet- improved visualization of the surgical field. It aids in tailor-
ic fields. Undoubtedly, high field, i.e. 1.5 or even 3 T systems ing craniotomies and localizing subcortical lesions. Yet, the
offer an excellent image quality that cannot be reached by usefulness of neuronavigation is limited once the tumor
low-field systems. They are, however, very expensive, and resection has begun. Surgically induced edema, csf loss,
are yet limited to a small number of centers worldwide. and tumor resection itself lead to anatomical alterations,
In ioMRI in general, there is a constant trade-off between the so-called ‘‘brain-shift’’ [33]. Consequently, neuronaviga-
signal-to-noise ratio, access to the patient, and usable field of tion, as it is based on preoperative imaging data, is not
view. The optimal design of a magnet with regard to homo- reliable in terms of determining residual tumor tissue intrao-
geneity would be a complete sphere without opening, which peratively.
is obviously impossible in both conventional ‘‘closed’’ and Therefore, ioMRI should be superior to neuronavigation
intraoperative ‘‘open’’ MRI systems [28]. To overcome this in tumor resection with the aid of frameless neuronavigation:
problem and to achieve an acceptable signal-to-noise ratio in it enables the surgeon to update anatomical data intraopera-
low field open MRI systems, longer acquisition times are tively. The intraoperative update of the neuronavigation
necessary. While weaker magnets are usually less expensive system helps to precisely localize and target remaining
and can allow near real-time imaging, they suffer from poorer tumor tissue [24, 30]. Bergsneider et al. [34] indicated the
image resolution. High-field systems on the other hand are superiority of ioMRI versus neuronavigation when they
more expensive and permit only interruptive scanning. found that a higher percentage of tumor resection was possi-
From an image-quality point of view, cylindrical supercon- ble with low-field strength ioMRI guidance versus neurona-
ductive systems bear significant advantages relative to static vigation alone.
magnetic field strength and homogeneity [29]. In the Signa SP Hence, in order to update neuro-anatomical information
scanner, the central segment of a cylindrical system was taken intraoperatively, intraoperative imaging techniques such as
out, allowing for direct access to the patient in the scanner at MRI or ultrasound are mandatory. In this respect, ultrasono-
the expense of decreased field strength at the imaging isocenter graphic imaging is not only more difficult to interpret but
compared to 1.5 T. Nonetheless, image quality is still almost might also impose severe limitations in regards of reliability
comparable to high field imaging, and sufficient for the of findings, even for experienced surgeons [35 37]. Conse-
delineation of both, contrast-enhancing and non contrast- quently, intraoperative imaging in glioma surgery is the
enhancing lesions such as high- and low-grade gliomas [27]. domain of MRI, allowing even for the incorporation of
In contrast to this system, e.g. the PoleStar ioMRI system preoperatively acquired high-resolution datasets. As out-
offers more convenience for the surgeon with a larger work- lined above, image quality issues might restrict the reliabili-
ing space, yet its field strength is even lower, and the mag- ty of ioMRI in non-contrast enhancing gliomas. Yet, no such
net’s field of view is considerably smaller, requiring limitations apply to ioMRI-guidance in enhancing lesions
accurate patient positioning and rendering a slightly lower amenable to surgical resection.
signal-to-noise ratio. Still, the PoleStar provides good quali-
ty visualization of contrast-enhancing tumors and fair quali-
ty for non-enhancing lesions [20, 30]. To our personal
experience, image quality in high-grade gliomas is compa- Influence of ioMRI on the Course of Surgery
rable in both systems, but image quality in low-grade glio-
mas is to some extent better when using the Signa SP [19]. While intraoperative magnetic resonance imaging has evol-
Contrast enhancing gliomas are displayed best on ved as a new technology in neurosurgery from the 1990s,
T1-weighted imaging with application of contrast agent, first reports dealing with ioMRI systems focused primarily
while usually higher doses of contrast agent are recom- on the feasibility of their use. Until today, more than 800
mended for optimal lesion-to-white-matter-contrast [31]. glioma patients have reportedly undergone surgical resec-
Several groups have experienced difficulties with visualiz- tion with the aid of any low field strength ioMRI. With the
ing the tumor margins on regular T2-weighted imaging in increasing experience of applying these methods in the neu-
non-enhancing tumors with different low field MRI units rosurgical routine and the knowledge of their safety, the
Table 1 Studies reporting the use of low field intraoperative MRI in influence of ioMRI on the surgical routine and clinical
low grade gliomas with a minimum of ten patients benefit of the patients has become more important.
Author Year System used No. of Increased In both, high- and low-grade glioma cases, all investiga-
patients resection due to
intraoperative tors who reported on the influence of intraoperative low field
imaging MRI on the course of surgery stated that intraoperative
Black et al. 1999 Signa SP 29 n.r. scanning had revealed residual tumor tissue in a significant
Seifert et al. 1999 Signa SP 13 n.r. number of cases. The percentage of patients with an extend-
Wirtz et al. 2000 Magnetom 29 45% ed resection after depiction of residual tumor-tissue on
Open intraoperative images is reported to range between 14 and
Zimmermann 2000 Signa SP 16 n.r.
et al.
75% (Table 1) [14, 15, 24, 30, 38 40]. Correspondingly, in
Bohinski et al. 2001 AIRIS II 10 40% high-grade glioma cases the respective figures are 10 71%
Zimmermann 2001 Signa SP 11 n.r. (Table 2) [14, 15, 22, 24, 30, 40 42]. As a consequence, the
et al. rates of complete resection have increased by usually greater
Buchfelder et al. 2002 Magnetom 11 14% than 20% [43]. Evidently, the use of low field ioMRI has had
Open
a major influence on the course of surgery in a large number
Nimsky et al. 2002 Magnetom 47 38%
Open of patients.
Open
Schulder et al. 2003 PoleStar N 10 12 n.r.
Claus et al. 2005 Signa SP 156 n.r. Influence of ioMRI on Patient Outcome
Open
Senft et al. 2008 PoleStar N 20 21a 47% In studies addressing postoperative morbidity, complication
Abbreviations: n.r. not reported rates lay between 5 and 18% directly postoperatively in all
a
Non enhancing gliomas but one series [15, 21, 27, 30, 32, 42, 44, 45], where an early
complication rate of 53% in a group of 13 patients was
Table 2 Studies reporting the use of low field intraoperative MRI in observed [34]. Permanent deficits occurred in constantly
high grade gliomas with a minimum of ten patients less than 10% of the patients, which is comparable to the
Author Year System used No. of Increased complication rates in conventional glioma surgery. In our
patients resection due to own series of patients, an extended resection after ioMRI
intraoperative scans had revealed residual tumor did not correlate with
imaging
postoperative morbidity [30]. To prevent patient injury and
Tronnier et al. 1997 Magnetom 10 n.r.
Open neurological deterioration due to overly aggressive resec-
Black et al. 1999 Signa SP 19 n.r. tion, the use of ioMRI can also be combined with intrao-
Knauth et al. 1999 Magnetom 41 41% perative monitoring techniques, such as motor or sensory
Open evoked potentials during surgery, even when applied in the
Seifert et al. 1999 Signa SP 12 n.r. fringe field [46]. As a result, there are no obstacles to
Wirtz et al. 2000 Magnetom 68 63%
conclude that, after careful patient selection, surgical resec-
Open
Zimmermann 2000 Signa SP 16 n.r. tion of gliomas can be increased to a maximum extent with
et al. acceptable risks for the patients, given the proven benefit of
Bohinski et al. 2001 AIRIS II 30 56% maximum resection.
Zimmermann 2001 Signa SP 14 n.r. Yet, only few studies have aimed to evaluate the influence
et al.
of intraoperative low field MRI on the clinical course of
Open glioma patients so far. As mentioned earlier, most studies
Trantakis et al. 2003 Signa SP 68 n.r. have focused on the feasibility, safety, and influence on the
Nimsky et al. 2003 Magnetom 54 13% course of surgery in the application of ioMRI in glioma
Open surgery. In a large series of low-grade glioma patients,
Schulder et al. 2003 PoleStar N 10 25 n.r. Claus et al. [2] found a trend towards prolonged survival of
Bergsneider et al. 2005 Magnetom 10 n.r.
patients who had undergone total vs. subtotal resection in an
Open
Hirschberg et al. 2005 Signa SP 32 52% ioMRI-environment. Similar findings were reported by Tran-
Schneider et al. 2005 Signa SP 31 71% takis et al. [47] in patients with high-grade gliomas. Using two
Senft et al. 2008 PoleStar N 20 42a 28% different low field systems, Wirtz et al. [40] and Schneider
Abbreviations: n.r. not reported et al. [42] independently reported statistically significant
a
Contrast enhancing gliomas prolonged survival times for high-grade glioma patients who
had undergone total vs. subtotal tumor resection with the aid of patients with brain tumors is obvious. The use of low field
of ioMRI. These results add to the growing evidence that ext- MRI scanners in glioma surgery is safe, reliable and most
ent of resection translates into prolonged survival in glioma useful in assessing the extent of resection of intrinsic, infil-
patients [4]. However, all theses studies did not compare trating brain tumors whose boundaries cannot be clearly
patients treated under ioMRI-guidance with a control group distinguished with the surgeon’s eyes. While several retro-
of patients. Hirschberg et al. [41] presented the only matched spective analyses have documented the benefit of imple-
group-analysis in a series of glioblastoma patients under- menting ioMRI into the neurosurgical routine in glioma
going intraoperative low-field ioMRI-guided vs. standard surgery by means of increasing the extent of resection,
microsurgery and found a trend in favor of ioMRI, but no prospective randomized studies are still lacking that might
statistically significant differences in overall survival between prove the benefit of ioMRI compared to the standard micro-
the groups. Therefore, the true benefit of ioMRI has not yet neurosurgical resection of brain tumors. Further, use of
been proven, in contrast to other means of intraoperative ioMRI itself has not been proven to prolong survival in
resection control, e.g. administration of fluorescent porphyr- patients with glial tumors. Although beneficial effects are
ins [5, 6]. Such studies are utterly needed today. suggested by previous studies, there is no high-class evi-
dence to promote the use of ioMRI today, and pursuant
studies are needed.
Comparison with High Field Systems One clear advantage over the administration of fluores-
cent porphyrins to visualize tumor tissue intraoperatively is
Even though the superiority of high field ioMRI systems the usability of ioMRI also in low-grade gliomas on the one
over low field ioMRI systems seems obvious in terms of hand, and the implementation of functional anatomical data
image resolution and quality, only few reports have specifi- on the other. In the future, not only pre-operative functional
cally addressed this issue. While Nimsky et al. [39] de- MRI datasets might be incorporated into neuronavigation,
scribed a clearly better image quality and a smoother but also intraoperative diffusion weighted imaging to depict
workflow with a high field system, the rates of extended white matter tracts seems feasible in both, high field and low
resection were comparable between the low and high field field ioMRI systems [48 50].
system, and Bergsneider et al. [34] found no difference in the
extent of resection when comparing a low field vs. a high
field ioMRI. High field systems might offer faster acquisi- Conclusions
tion times and might allow for a greater variety of imaging
sequences, but, as outlined before, low field systems, too, For obvious reasons, glioma patients will not be cured by
can provide the surgeon with reliable information about the neurosurgical intervention. Additional treatment modalities
extent of resection. To our own experience, this is definitely are and will remain mandatory to prolong survival and to
true at least for contrast enhancing, high-grade tumors. In maintain quality of life for these patients. However, maxi-
non-enhancing, low-grade tumors, high field systems surely mum tumor resection without induction of disabling neuro-
present an improved spatial resolution. logical deficits appears to be one of the strongest predictive
The question whether high and low field ioMRI differ in factors. Therefore, the neurosurgeon’s abilities and the
terms of patient benefit or clinical outcome remains none- advances of modern aides such as ioMRI will certainly trans-
theless unanswered. This is a crucial concern, however, late into benefit for our patients. In this respect, upcoming
especially when looking at the setup and installation costs studies will yet have to demonstrate the true value of ioMRI.
of the different available ioMRI systems. One distinct ad-
vantage of low field ioMRI systems over high field ones Conflict of interest statement T. Gasser serves as a clinical consul
certainly is the possibility to perform intraoperative electro- tant to Medtronic, but the company provided no payment or any other
benefits with respect to this work. V. Seifert and C. Senft declare to
physiological monitoring of eloquent function employing have no potentially conflicting interests.
sensory or motor evoked potential with the patient being in
the fringe field of the magnet without disturbances [46].
References
The Future of ioMRI in Glioma Surgery
1. Bennett AH, Godlee RJ (1974) A case of cerebral tumor the
surgical treatment. CA Cancer J Clin 24:171 181 (excerpt from
Advances in technology have increasingly improved the pre-
Trans R Med Chir Soc Lond 168:243 275, 1885)
cision of intracranial surgery, and ioMRI is the latest devel- 2. Claus EB, Horlacher A, Hsu L, Schwartz RB, Dello Iacono D,
opment in this field. The potential impact on the management Talos F, Jolesz FA, Black PM (2005) Survival rates in patients with
low grade glioma after intraoperative magnetic resonance image 18. Schulder M, Salas S, Brimacombe M, Fine P, Catrambone J,
guidance. Cancer 103:1227 1233 Maniker AH, Carmel PW (2006) Cranial surgery with an expanded
3. McGirt MJ, Chaichana KL, Gathinji M, Attenello FJ, Than K, compact intraoperative magnetic resonance imager. Technical
Olivi A, Weingart JD, Brem H, Quinones Hinojosa A (2009) note. J Neurosurg 104:611 617
Independent association of extent of resection with survival in 19. Ntoukas V, Krishnan R, Seifert V (2008) The new generation
patients with malignant brain astrocytoma. J Neurosurg 110:156 162 polestar n20 for conventional neurosurgical operating rooms: a
4. Sanai N, Berger MS (2008) Glioma extent of resection and its preliminary report. Neurosurgery 62:82 89, discussion 89 90
impact on patient outcome. Neurosurgery 62:753 764, discussion 20. Schulder M, Carmel PW (2003) Intraoperative magnetic resonance
264 756 imaging: impact on brain tumor surgery. Cancer Control 10:
5. Stummer W, Pichlmeier U, Meinel T, Wiestler OD, Zanella F, 115 124
Reulen HJ (2006) Fluorescence guided surgery with 5 aminolevu 21. Black PM, Alexander E 3rd, Martin C, Moriarty T, Nabavi A,
linic acid for resection of malignant glioma: a randomised con Wong TZ, Schwartz RB, Jolesz F (1999) Craniotomy for tumor
trolled multicentre phase III trial. Lancet Oncol 7:392 401 treatment in an intraoperative magnetic resonance imaging unit.
6. Stummer W, Reulen HJ, Meinel T, Pichlmeier U, Schumacher W, Neurosurgery 45:423 431, discussion 431 433
Tonn JC, Rohde V, Oppel F, Turowski B, Woiciechowsky C, 22. Knauth M, Wirtz CR, Tronnier VM, Aras N, Kunze S, Sartor K
Franz K, Pietsch T (2008) Extent of resection and survival in (1999) Intraoperative MR imaging increases the extent of tumor
glioblastoma multiforme: identification of and adjustment for bias. resection in patients with high grade gliomas. AJNR Am J Neuro
Neurosurgery 62:564 576, discussion 564 576 radiol 20:1642 1646
7. Albert FK, Forsting M, Sartor K, Adams HP, Kunze S (1994) Early 23. Metzger AK, Lewin JS (2001) Optimizing brain tumor resection.
postoperative magnetic resonance imaging after resection of Low field interventional MR imaging. Neuroimaging Clin N Am
malignant glioma: objective evaluation of residual tumor and its 11:651 657, ix
influence on regrowth and prognosis. Neurosurgery 34:45 60, 24. Nimsky C, Ganslandt O, Tomandl B, Buchfelder M, Fahlbusch R
discussion 60 61 (2002) Low field magnetic resonance imaging for intraopera
8. Bernays RL, Kollias SS, Khan N, Brandner S, Meier S, Yonekawa tive use in neurosurgery: a 5 year experience. Eur Radiol 12:
Y (2002) Histological yield, complications, and technological 2690 2703
considerations in 114 consecutive frameless stereotactic biopsy 25. Ram Z, Hadani M (2003) Intraoperative imaging MRI. Acta
procedures aided by open intraoperative magnetic resonance imag Neurochir Suppl 88:1 4
ing. J Neurosurg 97:354 362 26. Schulder M, Sernas TJ, Carmel PW (2003) Cranial surgery
9. Hall WA, Liu H, Martin AJ, Truwit CL (2000) Intraoperative and navigation with a compact intraoperative MRI system. Acta
magnetic resonance imaging. Top Magn Reson Imaging 11: Neurochir Suppl 85:79 86
203 212 27. Zimmermann M, Seifert V, Trantakis C, Kuhnel K, Raabe A,
10. Senft C, Seifert V, Hermann E, Gasser T (2009) Surgical treatment Schneider JP, Dietrich J, Schmidt F (2000) Open MRI guided
of cerebral abscess with the use of a mobile ultralow field MRI. microsurgery of intracranial tumours. Preliminary experience
Neurosurg Rev 32:77 85 using a vertical open MRI scanner. Acta Neurochir (Wien) 142:
11. Black PM, Moriarty T, Alexander E 3rd, Stieg P, Woodard EJ, 177 186
Gleason PL, Martin CH, Kikinis R, Schwartz RB, Jolesz FA (1997) 28. Hinks RS, Bronskill MJ, Kucharczyk W, Bernstein M, Collick BD,
Development and implementation of intraoperative magnetic reso Henkelman RM (1998) MR systems for image guided therapy.
nance imaging and its neurosurgical applications. Neurosurgery J Magn Reson Imaging 8:19 25
41:831 842, discussion 842 845 29. Lewin JS (1999) Interventional MR imaging: concepts, systems,
12. Seifert V, Zimmermann M, Trantakis C, Vitzthum HE, Kuhnel K, and applications in neuroradiology. AJNR Am J Neuroradiol
Raabe A, Bootz F, Schneider JP, Schmidt F, Dietrich J (1999) 20:735 748
Open MRI guided neurosurgery. Acta Neurochir (Wien) 141: 30. Senft C, Seifert V, Hermann E, Franz K, Gasser T (2008) Useful
455 464 ness of intraoperative ultralow field magnetic resonance imaging
13. Tronnier VM, Wirtz CR, Knauth M, Lenz G, Pastyr O, in glioma surgery. Neurosurgery 63:257 266, discussion 266 267
Bonsanto MM, Albert FK, Kuth R, Staubert A, Schlegel W, 31. Knauth M, Wirtz CR, Aras N, Sartor K (2001) Low field interven
Sartor K, Kunze S (1997) Intraoperative diagnostic and interven tional MRI in neurosurgery: finding the right dose of contrast
tional magnetic resonance imaging in neurosurgery. Neurosurgery medium. Neuroradiology 43:254 258
40:891 900, discussion 900 902 32. Nimsky C, Ganslandt O, Buchfelder M, Fahlbusch R (2003)
14. Wirtz CR, Bonsanto MM, Knauth M, Tronnier VM, Albert FK, Glioma surgery evaluated by intraoperative low field magnetic
Staubert A, Kunze S (1997) Intraoperative magnetic resonance resonance imaging. Acta Neurochir Suppl 85:55 63
imaging to update interactive navigation in neurosurgery: method 33. Roberts DW, Hartov A, Kennedy FE, Miga MI, Paulsen KD (1998)
and preliminary experience. Comput Aided Surg 2:172 179 Intraoperative brain shift and deformation: a quantitative analysis
15. Bohinski RJ, Kokkino AK, Warnick RE, Gaskill Shipley MF, of cortical displacement in 28 cases. Neurosurgery 43:749 758,
Kormos DW, Lukin RR, Tew JM Jr (2001) Glioma resection in a discussion 758 760
shared resource magnetic resonance operating room after optimal 34. Bergsneider M, Sehati N, Villablanca P, McArthur DL, Becker DP,
image guided frameless stereotactic resection. Neurosurgery Liau LM (2005) Mahaley Clinical Research Award: extent of
48:731 742, discussion 742 744 glioma resection using low field (0.2 T) versus high field (1.5 T)
16. Muragaki Y, Iseki H, Maruyama T, Kawamata T, Yamane F, intraoperative MRI and image guided frameless neuronavigation.
Nakamura R, Kubo O, Takakura K, Hori T (2006) Usefulness of Clin Neurosurg 52:389 399
intraoperative magnetic resonance imaging for glioma surgery. 35. Erdogan N, Tucer B, Mavili E, Menku A, Kurtsoy A (2005)
Acta Neurochir Suppl 98:67 75 Ultrasound guidance in intracranial tumor resection: correlation
17. Hadani M, Spiegelman R, Feldman Z, Berkenstadt H, Ram Z with postoperative magnetic resonance findings. Acta Radiol
(2001) Novel, compact, intraoperative magnetic resonance 46:743 749
imaging guided system for conventional neurosurgical operating 36. Rygh OM, Selbekk T, Torp SH, Lydersen S, Hernes TA, Unsgaard G
rooms. Neurosurgery 48:799 807, discussion 807 809 (2008) Comparison of navigated 3D ultrasound findings with
histopathology in subsequent phases of glioblastoma resection. 44. Bernstein M, Al Anazi AR, Kucharczyk W, Manninen P,
Acta Neurochir (Wien) 150:1031 1044 Bronskill M, Henkelman M (2000) Brain tumor surgery with the
37. Tronnier VM, Bonsanto MM, Staubert A, Knauth M, Kunze S, Toronto open magnetic resonance imaging system: preliminary
Wirtz CR (2001) Comparison of intraoperative MR imaging results for 36 patients and analysis of advantages, disadvantages, and
and 3D navigated ultrasonography in the detection and resection future prospects. Neurosurgery 46:900 907, discussion 907 909
control of lesions. Neurosurg Focus 10:E3 45. Zimmermann M, Seifert V, Trantakis C, Raabe A (2001) Open
38. Buchfelder M, Fahlbusch R, Ganslandt O, Stefan H, Nimsky C MRI guided microsurgery of intracranial tumours in or near elo
(2002) Use of intraoperative magnetic resonance imaging in tai quent brain areas. Acta Neurochir (Wien) 143:327 337
lored temporal lobe surgeries for epilepsy. Epilepsia 43:864 873 46. Szelenyi A, Gasser T, Seifert V (2008) Intraoperative neurophysio
39. Nimsky C, Ganslandt O, Fahlbusch R (2005) Comparing 0.2 tesla logical monitoring in an open low field magnetic resonance imag
with 1.5 tesla intraoperative magnetic resonance imaging analysis ing system: clinical experience and technical considerations.
of setup, workflow, and efficiency. Acad Radiol 12:1065 1079 Neurosurgery 63:268 275, discussion 275 276
40. Wirtz CR, Knauth M, Staubert A, Bonsanto MM, Sartor K, Kunze S, 47. Trantakis C, Winkler D, Lindner D, Strauss G, Nagel C,
Tronnier VM (2000) Clinical evaluation and follow up results Schneider JP, Meixensberger J (2003) Clinical results in MR guided
for intraoperative magnetic resonance imaging in neurosurgery. therapy for malignant gliomas. Acta Neurochir Suppl 85:65 71
Neurosurgery 46:1112 1120, discussion 1120 1122 48. Nimsky C, Ganslandt O, Hastreiter P, Wang R, Benner T, Sorensen
41. Hirschberg H, Samset E, Hol PK, Tillung T, Lote K (2005) Impact AG, Fahlbusch R (2007) Preoperative and intraoperative diffusion
of intraoperative MRI on the surgical results for high grade glio tensor imaging based fiber tracking in glioma surgery. Neurosur
mas. Minim Invasive Neurosurg 48:77 84 gery 61:178 185, discussion 186
42. Schneider JP, Trantakis C, Rubach M, Schulz T, Dietrich J, 49. Ozawa N, Muragaki Y, Nakamura R, Lseki H (2008) Intraopera
Winkler D, Renner C, Schober R, Geiger K, Brosteanu O, Zimmer tive diffusion weighted imaging for visualization of the pyramidal
C, Kahn T (2005) Intraoperative MRI to guide the resection of tracts. Part II: clinical study of usefulness and efficacy. Minim
primary supratentorial glioblastoma multiforme a quantitative Invasive Neurosurg 51:67 71
radiological analysis. Neuroradiology 47:489 500 50. Schulder M, Azmi H, Biswal B (2003) Functional magnetic reso
43. Oh DS, Black PM (2005) A low field intraoperative MRI system for nance imaging in a low field intraoperative scanner. Stereotact
glioma surgery: is it worthwhile? Neurosurg Clin N Am 16:135 141 Funct Neurosurg 80:125 131
Intraoperative MRI (ioMRI) in the Setting of Awake
Craniotomies for Supratentorial Glioma Resection
Pierpaolo Peruzzi, Erika Puente, Sergio Bergese, and E. Antonio Chiocca
Abstract Both awake craniotomy under conscious sedation After its initial introduction in 1994 at Brigham and
and use of intraoperative MRI can increase the efficiency Women’s Hospital in Boston, intraoperative MRI (ioMRI)
and safety of glioma resections. In contrast to craniotomies is being increasingly accepted as an adjunct in the surgical
under general anesthesia, neurosurgery under conscious treatment of cerebral gliomas [1 3]. One of the rationales for
sedation requires several changes to the routine operative its implementation in neuro-oncologic surgery derives from
setup when performed in the ioMRI environment. This work increasing evidence that extent of tumor resection correlates
reports our experience with awake craniotomies under con- positively with survival [4 7]. Consequently, the ability to
scious sedation using ioMRI. determine in the operating room if and how much tumor has
Seven patients underwent awake-craniotomies for resec- been left behind, allows the surgeon to decide whether to
tion of supratentorial gliomas using ioMRI at the Ohio State abstain from or continue with further resection.
University Medical Center and James Cancer Hospital by a Moreover, recent publications have discussed the ability
single surgeon. of this technology to aid in neuronavigation and reduce the
ioMRI can be safely employed in patients who are under- error associated with intraoperative brain shift [8 10].
going craniotomies under conscious sedation. Particularly Data are scarce on the use of ioMRI in the subpopulation
important is the evaluation by the anesthesiologist whether of patients harboring lesions located in eloquent areas that
the patient is a good candidate to sustain a likely longer than undergo surgery under conscious sedation. In comparison to
average procedure in a setting where his active cooperation craniotomies under general anesthesia, some may view the
is not only required, but also the essential aspect of this management of the awake patient during the scanning pro-
procedure. cedure as challenging. For some ioMRI solutions, such
as the PoleStar (Medtronic, Inc., Minneapolis, MN), the
Keywords Conscious sedation Craniotomy Glioma patients is inside a Faraday copper-wire tent during the
Intraoperative MRI scan and thus visibility of the patient under conscious seda-
tion is curtailed. Even with high-strength ioMRIs, the
Abbreviation patient’s face is hidden from view and thus lack of its
visibility could be an issue in ensuring that adequate venti-
ioMRI intraoperative Magnetic Resonance Imaging
lation and patient comfort are not being affected.
We thus analyzed retrospectively whether we could
perform ioMRI in patients undergoing surgical resections
of gliomas under conscious sedation (CS). Over a total of
40 patients who underwent craniotomy under CS for resec-
tion of supratentorial gliomas we identified 7 cases (17%)
where ioMRI was performed. The time frame of the analy-
P. Peruzzi and E.A. Chiocca (*) sis was from January 2005 to June 2009 and all cases were
Department of Neurological Surgery, The Ohio State University performed at the Ohio State University Medical Center and
College of Medicine and James Cancer Hospital, N 1017 Doan Hall, James Cancer Hospital by a single surgeon (EAC).
410 West Tenth Avenue, Columbus, OH, 43210, USA
Here we describe three such cases with the aim of
e mail: EA.Chiocca@osumc.edu
providing practical insights on the use of ioMRI in the
E. Puente and S. Bergese
Department of Anesthesiology, The Ohio State University College of consciously sedated patient undergoing glioma resections
Medicine and James Cancer Hospital, Columbus, OH, USA in eloquent areas.
44 P. Peruzzi et al.
Case Reports localized in the gyrus posterior to the one that contained the
lesion. Tumor debulking proceeded uneventfully testing the
Case 1 patient in the performance of requested motor and speech
tasks. When the operating surgeon estimated that gross total
This 40-year old right-handed male, in otherwise good resection of the tumor had occurred, another ioMRI was
health, presented to the emergency room (ER) with the first obtained. Additional diprivan was administered to sedate
time occurrence of a generalized tonic-clonic (GTC) seizure. the patient and then the copper-wired tent (Starshield, Med-
Brain Magnetic Resonance Imaging (MRI) showed a non tronic, Inc.) was placed on top and around the patient. After
enhancing right frontal lesion in proximity to the primary scanning for a period of approximately 16 min (T2 axial and
motor area. His neurologic exam, upon resolution of the ictal FLAIR axial, 7 and 9 min each), the ioMRI suggested a gross
period, was unremarkable. The imaging features suggested a total resection (Fig. 1, Ib) and no further removal was pur-
low grade glioma, primarily confined within gyri in the sued. Importantly, the scanning process was well tolerated
premotor strip and it was thus decided to try and attempt a by the patient and no technical difficulties were encountered.
radical resection under CS. Postoperatively the patient had a slight decrease in
After conscious sedation was obtained with diprivan and strength in his left upper extremity, in the order of 5-/5
dexmedetomidine in the operating room, a baseline ioMRI (Medical Research Council, MRC scale). However, this
scan was obtained (Fig. 1, Ia). Thereafter, a right frontal did not impair his functional status. A postoperative MRI
craniotomy was performed and once the brain was exposed, performed within 24 h after surgery confirmed gross total
cortical stimulation was used to localize motor cortex. As resection of the tumor (Fig. 1, Ic Id). The patient was dis-
expected from the preoperative MRI, the motor cortex was charged to home in stable conditions on postoperative
Fig. 1 Upper row: T2 weighted ioMR imaging obtained prior to tumor resection shows hyperintense signal in the right frontal lobe (Ia). Repeated
scan after resection shows surgical cavity and removal of lesion (Ib), as confirmed by postoperative MRI, respectively axial FLAIR (Ic) and
coronal T1 weighted with contrast (Id). Middle row: pre resection ioMRI shows a large left frontal lesion hyperintense in T2 weighted sequences
(IIa). Follow up intraoperative scan shows residual tumor as a hyperintense area in the posteromedial part of the surgical cavity (IIb) which was
thereafter addressed. Postoperative axial (IIc) and coronal (IId) T1 weighted images with constrast show gross total tumor resection. Lower row:
Preoperative MRI is significant for a large left frontotemporal lesion (IIIa) with areas of contrast enhancement (IIIb). Postoperative imaging shows
large surgical cavity without evidence of residual tumor (IIIc IIId). In this case ioMRI was not obtained because of technical difficulties
Intraoperative MRI (ioMRI) in the Setting of Awake Craniotomies for Supratentorial Glioma Resection 45
day 4 with a diagnosis of WHO grade 2 oligodendroglioma, Case 3

for which he subsequently received chemotherapy with temozo-
lomide, and is currently neurologically stable 18 months after
This 18-year old left-handed male was referred to our clinics
surgery without clinical or radiological evidence of tumor
to be evaluated for a newly discovered FLAIR-hyperintense
progression.
abnormality involving his right insular region (Fig. 1, IIIa),
highly suspiscious for a low grade glioma. Apart from head-
aches he was neurologically asymptomatic. Surgical removal
of the lesion was recommended both for diagnostic and prog-
Case 2 nostic purposes, and craniotomy under conscious sedation
was planned in order to minimize surgical morbidity. Once
in the operating room, after sedation was achieved, an initial
This 44-year old right-handed female with a history of a
ioMRI was obtained for baseline (Fig. 1, IIIb). The initial part
left frontal WHO grade 2 oligoastrocytoma, partially resected
of the craniotomy was uneventful, but when the patient was
8 and 4 years prior to the current admission, was referred to
allowed to regain consciousness in order to proceed with the
our institution with increased frequency of seizure activity,
cortical mapping and tumor removal, he became progressive-
manifesting as a right sided jacksonian progression, mainly
ly agitated and restless, to the point that induction of general
involving her lower extremity. Her most recent brain MRI
anesthesia was necessary. It was decided, however, to continue
was consistent with tumor recurrence at the site of her left
with resection and the tumor was debulked until no clear
motor area and suggestive of likely malignant progression.
neoplastic tissue was evident under the microscope. However,
Since it was felt that histological analysis would be necessary
a frozen section of the cavity margin showed presence of
to guide further treatment and to improve seizure control, a
scattered atypical cells dispersed within normal white matter.
re-resection of her recurrent lesion was attempted.
At this point an ioMRI was obtained, showing residual T2
Once in the operating room, conscious sedation was
hyperintensity in the more medial aspect of the tumor (Fig. 1,
induced using diprivan and dexmedetomidine and then a
IIIc). Further resection was then attempted, but shortly there-
baseline ioMRI was obtained inside the copper-wire tent
after perforating vessels feeding the basal ganglia were
(Fig. 1, IIa). Cortical mapping for speech and motor cortex
encountered and we desisted from further resection. Posto-
was performed with an Ojemann stimulator. During the
peratively, his neurological status remained intact. MR imag-
surgical resection of the tumor, the patient was also periodi-
ing showed a subtotal resection of the tumor mass, confirming
cally asked to move her right side and to talk with a speech
the persistence of the medial and posterior portion of the
pathologist present in the room. Once the lesion was judged
lesion (Fig. 1, IIId). The patient was discharged to home
to have been removed in a gross total fashion, a second
2 days after surgery with a final diagnosis of low grade
ioMRI was obtained. This showed some residual tumor in
astrocytoma.
the most medial part of the hemisphere, along the falx
(Fig. 1, IIb). Further resection was then performed until the
tumor was judged to have been resected in a gross total
fashion.
In this second phase of the resection, the patient did not Discussion
show intraoperative signs of neurological impairment. She
was able to tolerate ioMRI well and no technical difficulties In the past few decades, improvements in the treatment
ensued. Postoperative MRI corroborated the preliminary of many neurological diseases have been tied to technologi-
findings of the gross total resection (Fig. 1, IIc IId); Histo- cal advances. Specifically, for the treatment of gliomas the
logic evaluation upgraded her tumor to oligoastrocytoma introduction of intraoperative MRI, together with the use of
with isolated areas of anaplastic progression. stereotactic navigation, can be useful adjuncts to achieve the
Postoperatively, however, her neurological status was therapeutic goal of gross total resection. In fact, the extent of
overall worsened as compared to her preoperative function, tumor removal appears to influence disease progression with
with new severe right sided weakness and impairment of fine numerous reports showing that the incidence of tumor recur-
motor skills most consistent with a supplementary motor rence and malignant transformation of low grade gliomas are
area (SMA) syndrome. Eventually she needed 2 weeks of significantly reduced [7, 8, 11, 12], while extended survival
inpatient rehabilitation before being able to be discharged to is achieved for higher grade lesions [5, 6]. In 40 70% of
home and to date, 6 months after surgery, she has recovered patients, the adjunct of ioMRI has been reported to lead to
most of her motor functions and functional independence, discovery of residual tumor [2, 13 15], with a greater than
seizure activity has been well controlled since surgery and 20% increase in cases of total tumor resection [13, 16 18].
she is still tumor free. This advantage seems to be more prominent for low grade
gliomas compared to high grade gliomas which often have compatible with the need of the neurosurgeon to have the
signs of necrosis or grossly abnormal appearance [16, 18]. patient able to talk and to follow commands.
The predictive value of ioMRI for reliable detection of The main challenge for the anesthesiologist in awake
tumor residues has been validated even with the low-field craniotomy procedures is to provide adequate sedation and
strength units. Hirschl et al. have recently shown a 82% analgesia during the different stages of the procedure [21,
concordance between intraoperative and post-operative 22]. The primary goal is to keep the patient awake and alert
imaging; more importantly, they reported ‘‘false positive’’ during brain mapping, allowing for full cooperation, while
results, i.e. residual tumor detected by intraoperative images providing sufficient, safe and effective analgesia and guar-
while not identified in postoperative standard MR images, anteeing comfort to the patient. It is paramount that the
only in 1% of cases [10]. The question remains whether anesthesiologist evaluate patients who may benefit from
ioMRi should be applied to all craniotomies for glioma this procedure thoroughly, since some patients may not be
resection or if it should be dedicated only to selected cases. ideal candidates to undergo awake procedures. During this
In our experience, we reserve ioMRI use for low grade phase, the assessment of the patients’ level of comfort and
tumors since its more normal appearance can escape intrao- cooperation with the described procedure is made. Also,
perative detection. possible contraindications to the procedure, like esophageal
With lesions involving eloquent areas, ioMRI can also be reflux, obesity and large vascular tumors are ruled out.
performed with patients undergoing ‘‘awake’’ craniotomies. Patients should be counseled before the surgery and optimally
Little data is available in the literature reporting results of have the opportunity to interact with other patients that have
the combination of the two procedures. In their description gone through this procedure. We have observed that some-
of a series of 25 patients who underwent ioMRI for resection times the awake technique in patients of younger age, like
of low grade gliomas, Martin et al. included at least five the one reported in our Case 3, and generally those with
patients (the exact number is not specified by the authors) higher sympathetic tone is not successful. In our case, several
who, in addition to intraoperative imaging were operated attempts where made to adjust the anesthetic regimen, such
under conscious sedation because of the location of their as, increasing the level of a-2 agonists (dexedetomidine) and
lesions. In these five patients, despite the evidence of tumor decreasing GABA receptor agonist drugs (like propofol or
remnant at ioMRI, total resection could not be achieved midazolam) with the intention of obtaining a higher degree
because of the onset of changes in neurological functions, of cooperation. However, in this patient, the fine balance
and four experienced mild transient deficits in the immediate between an adequate level of sedation and good cooperation
postoperative period, while the other one suffered a was lost, and that resulted in oversedation with resultant loss
permanent left-sided paresis [7]. of a meaningful cooperation. In general, when such situa-
Another such case is reported by Tan and Leong, where tions happen, after discussing with the surgeons, it usually
a middle aged woman was diagnosed with a low grade becomes necessary to convert to conventional general anes-
glioma which was completely resected without major thesia techniques to achieve proper anesthesia levels.
complications. The authors, though, stress the fact that the The most popular techniques that are utilized are based on
procedure was unduly lengthy because of the multiple a combination of general anesthesia and an awake technique,
intraoperative scans, leading to patient exhaustion during better known as ‘‘asleep awake asleep’’ technique (AAA).
the procedure [19]. This technique places patients’ under general anesthesia
Awake craniotomies in the context of ioMRI were before and after brain mapping. Therefore, it consists of an
reported in 38 cases performed from 2005 to 2008 at the initial phase of general anesthesia, followed by an intrao-
university of Kiel, Germany, where the authors mainly perative awakening phase and finally, back to general anes-
stressed the ‘‘emotional’’ aspect of the procedure, conclud- thesia [23, 24]. The recent introduction of dexmedetomidine,
ing that, despite being physically and psychologically a highly selective a2-agonist, with dose-dependent sedative,
demanding for the patients, it was overall well tolerated, anxyolitic, and analgesic effects has provided an alternative
leading them to adopt a standard approach by which all to sedation that decreases respiratory suppression and
awake craniotomies for gliomas in their institution are per- reduces opioid administration. Dexmedetomidine allows
formed with ioMRI implementation [20]. minimal sedation, resulting in an awake and cooperative
The role of the anesthesiologist is crucial both preopera- state that enables patient’s to feel the effect of sedation and
tively, when it comes to determine if each single patient is a at the same time respond to cognitive tests, necessary to
suitable candidate, medically and emotionally, for this assess specific functions that may be impaired during
stressful procedure, and, intraoperatively, with regards of tumor resection. The addition of ioMRI to awake craniotomy
maintaining an appropriate level of anesthesia and analgesia increases the complexity of the whole procedure for several
Intraoperative MRI (ioMRI) in the Setting of Awake Craniotomies for Supratentorial Glioma Resection 47
aspects. The space constriction and the noise which the Conclusion
patient is subjected to during the scanning phase may not
be well tolerated and may require anesthesia adjustments, ioMRI can be safely employed in patients who are under-
increasing the Awake-Asleep phases needed throughout the going craniotomies under conscious sedation. Particularly
case. Moreover, during the time the patient is inside of the important, in this setting, is the evaluation by the anesthesi-
Faraday cage, within the bore of the scanner, he will not be ologist whether the patient is a good candidate to sustain a
under the direct supervision or visibility of the anesthesiolo- likely longer than average procedure in a setting where his
gist for the duration of the scan. During this period, it is active cooperation is not only required, but also the essential
essential to provide continuous ventilatory support, provide aspect of this procedure.
special attention to IV lines, drugs or contrast media infu-
sions and inhalation anesthetics, and constantly supervise Conflict of interest statement We declare that we have no conflict
of interest.
monitoring devices, in compliance with the American Soci-
ety of Anesthesiologists (ASA) guidelines.
Due to the difficulties in monitoring and limited access to
the patient, while in the cage, complications should be References
expected such as, inadequate or excessive sedation, pain,
nausea or vomiting; airway management difficulties or 1. Black PM, Moriarty T, Alexander ER, Stieg P, Woodard EJ,
obstruction, decrease in oxygen saturation, below 90%, Gleason PL, Martin CH, Kikinis R, Schwartz RB, Jolesz FA
hypoventilation, below 8 bpm, hemodynamic instability (1997) Development and implementation of intraoperative mag
and new neurological symptoms and signs [13]. netic resonance imaging and its neurosurgical applications. Neuro
surgery 41:831 842, discussion 842 845
No statistically relevant data can be inferred from our 2. Maesawa S, Fujii M, Nakahara N, Watanabe T, Saito K, Kajita Y,
experience of using ioMRI in the setting of awake cranio- Nagatani T, Wakabayashi T, Yoshida J (2009) Clinical indications
tomies. However, we believe that our case series is repre- for high field 1.5 T intraoperative magnetic resonance imaging and
sentative of some of the most common situations neuro navigation for neurosurgical procedures. Review of initial
100 cases. Neurol Med Chir (Tokyo) 49:340 349, discussion
encountered in the operating room and give us the opportu- 349 350
nity to discuss our approach to this complex operative 3. Schwartz RB, Hsu L, Wong TZ, Kacher DF, Zamani AA, Black PM,
setting. Case 1 describes a situation where the ioMRI is Alexander ER, Stieg PE, Moriarty TM, Martin CA, Kikinis R,
mainly used as a confirmatory test of gross total resection. Jolesz FA (1999) Intraoperative MR imaging guidance for intra
cranial neurosurgery: experience with the first 200 cases. Radiolo
In our experience, the decision to use ioMRI is usually made gy 211:477 488
‘‘a priori’’, particularly in an awake case, since we believe 4. Albert FK, Forsting M, Sartor K, Adams HP, Kunze S (1994)
that the approach of ‘‘scanning everybody whenever possi- Early postoperative magnetic resonance imaging after resection of
ble’’, like other authors suggest [20, 25], is not necessary. malignant glioma: objective evaluation of residual tumor and its
influence on regrowth and prognosis. Neurosurgery 34:45 60,
Instead, the decision of performing ioMRI should be dictated discussion 60 61
by different variables, like tumor location, radiological 5. Hess KR (1999) Extent of resection as a prognostic variable in the
aspect, size, and surgical accessibility to the tumor. Case 2 treatment of gliomas. J Neurooncol 42:227 231
portrays a situation where ioMRI proved valuable in deter- 6. Lacroix M, Abi Said D, Fourney DR, Gokaslan ZL, Shi W,
DeMonte F, Lang FF, McCutcheon IE, Hassenbusch SJ, Holland E,
mining surgeon’s decision to carry on with tumor resection, Hess K, Michael C, Miller D, Sawaya R (2001) A multivariate
eventually helping to achieve a gross total resection. This is analysis of 416 patients with glioblastoma multiforme: prognosis,
the situation where a fine balance between extent of resec- extent of resection, and survival. J Neurosurg 95:190 198
tion and preservation of neurological function has to be 7. Martin C, Alexander ER, Wong T, Schwartz R, Jolesz F, Black PM
(1998) Surgical treatment of low grade gliomas in the intraopera
maintained. Our patient did experience some postoperative tive magnetic resonance imager. Neurosurg Focus 4:e8
deficits despite a substantially uneventful procedure, as it 8. Hall WA, Kowalik K, Liu H, Truwit CL, Kucharezyk J (2003)
has been described multiple times in the setting of awake Costs and benefits of intraoperative MR guided brain tumor resec
craniotomies and most commonly secondary to postopera- tion. Acta Neurochir Suppl 85:137 142
9. Hall WA, Liu H, Maxwell RE, Truwit CL (2003) Influence of
tive brain swelling or postoperative regional ischemia [26] 1.5 Tesla intraoperative MR imaging on surgical decision making.
but these resolved on postoperative followup. Acta Neurochir Suppl 85:29 37
Lastly, our case 3 reminds how a balance between proper 10. Hirschl RA, Wilson J, Miller B, Bergese S, Chiocca EA (2009)
sedation and ability to raise consciousness needs to be The predictive value of low field strength magnetic resonance
imaging for intraoperative residual tumor detection. Clinical arti
achieved or else patients will suffer undue agitation, partic- cle. J Neurosurg 111:252 257
ularly as they awaken, leading to inability to implement the 11. Albert FK, Forsting M (2003) Resection and prognosis. J Neuro
preoperative plan of action. surg 98:225 226, author reply 226
12. Berger MS, Deliganis AV, Dobbins J, Keles GE (1994) The effect 19. Tan TK, Leong KW (2009) Awake craniotomy in an intra
of extent of resection on recurrence in patients with low grade operative MRI environment. Anaesthesia 64:575 576
cerebral hemisphere gliomas. Cancer 74:1784 1791 20. Nabavi A, Goebel S, Doerner L, Warneke N, Ulmer S, Mehdorn M
13. Senft C, Seifert V, Hermann E, Franz K, Gasser T (2008) Usefulness (2009) Awake craniotomy and intraoperative magnetic resonance
of intraoperative ultra low field magnetic resonance imaging in glio imaging: patient selection, preparation, and technique. Top Magn
ma surgery. Neurosurgery 63:257 266, discussion 266 267 Reson Imaging 19:191 196
14. Staubert A, Pastyr O, Echner G, Oppelt A, Vetter T, Schlegel W, 21. Bergese SD, Puente EG (2009) Anesthesia in the intraoperative
Bonsanto MM, Tronnier VM, Kunze S, Wirtz CR (2000) An MRI environment. Neurosurg Clin N Am 20:155 162
integrated head holder/coil for intraoperative MRI in open neuro 22. Zorzi F, Saltarini M, Bonasin P (2008) Anesthetic management in
surgery. J Magn Reson Imaging 11:564 567 awake craniotomy. Signa vitae 3:28 32
15. Wirtz CR, Tronnier VM, Bonsanto MM, Knauth M, Staubert A, 23. Conte V, Baratta P, Tomaselli P, Songa V, Magni L, Stocchetti N
Albert FK, Kunze S (1997) Image guided neurosurgery with (2008) Awake neurosurgery: an update. Minerva Anestesiol
intraoperative MRI: update of frameless stereotaxy and radicality 74:289 292
control. Stereotact Funct Neurosurg 68:39 43 24. Piccioni F, Fanzio M (2008) Management of anesthesia in awake
16. Nimsky C, Ganslandt O, Tomandl B, Buchfelder M, Fahlbusch R craniotomy. Minerva Anestesiol 74:393 408
(2002) Low field magnetic resonance imaging for intraoperative use 25. Whittle IR, Borthwick S, Haq N (2003) Brain dysfunction follow
in neurosurgery: a 5 year experience. Eur Radiol 12:2690 2703 ing ‘awake’ craniotomy, brain mapping and resection of glioma. Br
17. Oh DS, Black PM (2005) A low field intraoperative MRI system J Neurosurg 17:130 137
for glioma surgery: is it worthwhile? Neurosurg Clin N Am 26. Duffau H, Lopes M, Denvil D, Capelle L (2001) Delayed onset
16:135 141 of the supplementary motor area syndrome after surgical resection
18. Schulder M, Carmel PW (2003) Intraoperative magnetic resonance of the mesial frontal lobe: a time course study using intraoperative
imaging: impact on brain tumor surgery. Cancer Control 10:115 124 mapping in an awake patient. Stereotact Funct Neurosurg 76:74 82
Glioma Extent of Resection and Ultra-Low-Field ioMRI: Interim
Analysis of a Prospective Randomized Trial
Christian Senft, Andrea Bink, Michael Heckelmann, Thomas Gasser, and Volker Seifert
Abstract Aiming at providing high-class evidence regarding Introduction

the use of intraoperative MRI (ioMRI), we are conducting a
prospective randomized controlled trial. Adult patients with Intraoperative magnetic resonance imaging (ioMRI) has
contrast enhancing lesions suspicious of malignant gliomas been used as a surgical adjunct in the resection of brain
scheduled to undergo radiologically complete tumor resection tumors for more than a decade [1]. Many groups reported
are eligible to enter this trial. After giving their informed that ioMRI-guidance is beneficial in detecting unintention-
consent, patients are randomized to undergo either ioMRI- ally remaining tumor tissue intraoperatively leading to ex-
guided or conventional microneurosurgical tumor resection. tended tumor resections, irrespective of the field strength of
To assess the extent of resection, pre- and early postoperative the magnet [2 7]. Some studies have even demonstrated
high-field MR images are obtained to perform volumetric benefits in terms of survival when looking at the extent of
analyses. Primary endpoint of the study is the rate of radiolog- resection as a prognostic factor [8, 9].
ically complete tumor resections. After the inclusion of 35 But these reports mainly represent retrospective cohort
patients, we performed an interim analysis. In six patients, analyses, and only Hirschberg et al. [10] performed a
histopathological examination revealed metastases, so they matched-group analysis in an attempt to show superiority
were excluded from further analyses. Thus, data from 29 of ioMRI-guidance compared to conventional microneuro-
patients with gliomas could be analyzed. There were no sig- surgery. Using a low field strength system (Signa SP, 0.5 T)
nificant differences in patient age (P = 0.28) or preoperative this study could not show differences between the ioMRI and
tumor sizes (P = 0.40) between the two treatment groups. We the control group. An explanation for this result is the fact
observed a trend towards a higher rate of complete tumor that the extent of tumor resections was comparable in both
resections in the ioMRI-group compared to the control group groups. This was unsurprisingly so, as one would expect the
(P = 0.07). Postoperative tumor volumes were significantly extent of resection to be the prognostic factor, and ioMRI
lower in the ioMRI-group than in the control group (P < 0.05). might just be able to influence the extent of resection. There-
The use of ioMRI appears to be associated with a higher rate fore, prospective randomized trials are mandatory to deter-
of radiographically complete as well as near total tumor resec- mine beneficial effects of ioMRI-guided brain tumor surgery.
tions compared to conventional microneurosurgery. Here, on the occasion of the second Meeting of the
Intraoperative Imaging Society, we present the results of
Keywords Extent of resection Glioma surgery Intraopera- an interim analysis of such a trial. To our knowledge, no
tive MRI PoleStar other such study has been conducted so far.
C. Senft (*), M. Heckelmann, and V. Seifert

Schleusenweg 2 16, 60528, Frankfurt, Germany Material and Methods
e mail: c.senft@med.uni frankfurt.de
A. Bink
Department of Neuroradiology, Johann Wolfgang Goethe University, Patients
Frankfurt, Germany
T. Gasser
Adult patients with contrast enhancing lesions on T1-weighted
Schleusenweg 2 16, 60528, Frankfurt, Germany MRI suspicious of malignant gliomas who are scheduled
Department of Neurosurgery, University of Essen, Essen, Germany to undergo radiologically complete tumor resections are
50 C. Senft et al.
eligible to enter this trial. Also, patients with known gliomas positioning, overall survival, and neurological deficits. All
and contrast enhancing lesions suspicious of tumor recur- patients undergo high field MRI within 7 days prior to
rence may be included. Patients in which the tumor is located surgery as well as within 72 h after surgery. One experienced
close to eloquent structures, e.g. motor cortex, corticospinal neuroradiologist (A.B.) who is blinded for the surgical treat-
tract, basal ganglia, Broca’s or Wernicke’s area, in a way ment of the patients performs the analysis of high field
that the surgical goal would be a near total or partial resec- MRI. Volumetric analyses are performed for both, pre- and
tion to spare these structures are not included. The study is postoperative imaging studies. As done previously [12],
conducted at the Johann Wolfgang Goethe-University, residual tumor is defined as contrast-enhancement with a
Frankfurt am Main/Germany, in adherence to the guidelines volume >0.175 cm3 on postoperative MRI.
of the International Conference on Harmonisation Good
Clinical Practice. This study was approved by the local ethics
committee.
Statistics
Treatment Statistical analyses were performed using commercially

available software (BiAS for Windows 9.01, epsilon Verlag,
After having given written informed consent, patients are Frankfurt/Germany). Nominal dichotomized data were com-
randomized to undergo either ioMRI-guided or conventional pared with Fisher’s exact test (number of complete resec-
microneurosurgical resection. Intraoperative MRI-guided tions, patient sex). After testing for Gaussian distribution
neurosurgery implies the use of a mobile intraoperative using Kolmogorov-Smirnov’s test, Student’s T-tests were
ultra-low field ioMRI system (PoleStar-N 20, Odin Medical applied to compare patients’ age and the duration of sur-
Technologies, Yokneam, Israel/Medtronic, Louisville, CO, geries. If a Gaussian distribution could not be confirmed,
USA). The system can be used for intraoperative image Wilcoxon-Mann-Whitney-U-tests were used to test for dif-
acquisition for neuronavigation with its integrated software ferences in pre- and postoperative tumor volumes as well as
(StealthStation, Medtronic, Louisville, CO, USA) as well the duration of patient positioning. P values lower than 0.05
as intraoperative resection control as described previously were considered statistically significant.
[7, 11]. All intraoperative imaging is conducted and evalu-
ated by neurosurgeons. Conventional microneurosurgical
resection implies the use of all standard neurosurgical instru-
ments and techniques, e.g. the use of an ultrasonic aspirator Results
as well as a neuronavigation system. Two neuronavigation
systems are available at our department (StealthStation and
VectorVision, BrainLAB, Heimstetten, Germany). Tools Patient Demographics
that are capable to detect or visualize tumor tissue intrao-
peratively, e.g. intraoperative ultrasound or fluorescence- Between October 2007 and March 2009, a total of 35
guided surgery as described by Stummer et al. [12] are not Patients entered the study. 18 patients were randomized to
allowed to be used in either group. Following surgery, the undergo ioMRI-guided surgery (ioMRI-group), 17 were ran-
resected tissue is sent for independent histopathological domized to undergo conventional microsurgical tumor re-
analysis. Postsurgical treatment is conducted according to section (control group). Histopathology revealed metastasis
standard protocols and clinical guidelines, depending on in six patients (three in each group) who were excluded from
tumor histology, previous treatment and patient preferences. further analysis. With 17 male and 12 female patients over-
All patients are regularly followed up with clinical and MRI all, there was no sex imbalance between the groups (ioMRI-
examinations every 3 months. group: nine males, six females; control group: seven males,
six females; P = 1.00). Mean age was 56.2 years (range:
35 84 years). There were no age differences between the
ioMRI-group and the control group (mean: 53.9 vs. 58.6
Study Endpoints years, respectively; P = 0.28). All patients were in good
clinical condition with a median preoperative KPS score of
The primary endpoint of this study is the number of patients 90 (range: 60 100). There were no differences between the
without contrast enhancing tumor as determined by high ioMRI- and the control group (median KPS score 90 in both,
field MRI. Secondary endpoints are the volume of residual P = 0.40). Preoperative tumor volumes were not statistically
tumor on postoperative MRI, duration of surgery and patient different between the groups (P = 0.40, Fig. 1).
Glioma Extent of Resection and Ultra-Low-Field ioMRI: Interim Analysis of a Prospective Randomized Trial 51
Fig. 1 Box Whisker plot showing preoperative tumor volumes for both groups
Extent of Resection Table 1 Comparison of the number of patients in which complete

resections were achieved
Complete resection (no.) Incomplete resection (no.)
In the ioMRI-group, complete tumor resection was achieved ioMRI group 14 1
in 93.3% of the cases, while intraoperative imaging had led Control group 9 5
to extended tumor resection in 4 out of 15 patients (26.7%).
In the control group, complete tumor resection was achieved
64.3% of the cases. This difference did not quite reach statisti-
cal significance (P = 0.0695, Table 1). Median postoperative
Discussion
tumor volumes were 0.0 cm3 in the ioMRI group and
Previously, a number of retrospective cohort series have
0.065 cm3 in the control group. This difference was statisti-
indicated that ioMRI-guidance influences intraoperative de-
cally significant (P = 0.046, Fig. 2).
cision-making by frequently revealing remaining tumor tis-
sue [2, 7, 13]. The intraoperative image acquisition along
with continued tumor resection has led to lower tumor
volumes [4, 14]. Yet, it can be argued that not all imaging
Time Consumption sequences are performed when the surgeon believed to have
resected the intended amount of tumor. Such an objection
As for the special requirements of positioning the patient might fall short, as the possibility to update the neuronaviga-
within the scanner and its field of view, patient positioning tion system at any stage during the surgery might improve
took slightly longer in the ioMRI-group than in the control the safety of brain tumor surgery [15]. On the other hand, we
group. The median positioning times were 25.0 and need scientific evidence to not only claim benefits of intrao-
12.0 min, respectively. This difference did not reach statisti- perative image-guidance by means of ioMRI but to justify
cal significance (P = 0.122). Similarly, the duration of the the use of such expensive technology [16]. With that goal in
surgical procedure itself as measured from skin incision to mind, we are conducting the first randomized controlled trial
wound closure was longer in the ioMRI-group (mean: comparing ioMRI-guided resection to conventional micro-
248 min) than in the control group (mean: 232 min). surgical resection in glioma surgery.
Again, this difference did not reach statistical significance Our interim analysis with a yet small number of patients
(P = 0.569). indicates such beneficial effects. The study population was
52 C. Senft et al.
Fig. 2 Box Whisker plot

showing postoperative tumor
volumes for both treatment
groups
homogenous, patient age and preoperative tumor size were postoperative tumor volumes and might lead to higher
distributed evenly between the two groups. Apparently rates of complete resections than conventional microneuro-
due to small sample size, we could not yet see a statistically surgical approaches.
significant benefit of ioMRI-guidance over conventional
microsurgery in terms of complete tumor resection. There Conflict of interest statement T. Gasser serves as a clinical con
sultant to Medtronic, but the company provided no financial support or
is however, a clear trend towards a superiority of ioMRI- any other benefits with respect to this work. All other authors declare to
guidance. Accordingly, the postoperative tumor volumes as have no potentially conflicting interests.
determined by high-field MRI were lower in patients treated
with ioMRI-guidance. This advantage of ioMRI seems to be
at the cost of longer patient positioning and a somewhat References
longer duration of surgery, even if not reaching statistical
significance. 1. Black PM, Alexander E 3rd, Martin C, Moriarty T, Nabavi A,
The above are only preliminary conclusions from this in- Wong TZ, Schwartz RB, Jolesz F (1999) Craniotomy for tumor
terim analysis. However, in line with previous reports [17], treatment in an intraoperative magnetic resonance imaging unit.
Neurosurgery 45:423 431, discussion 431 433
ioMRI led to extended tumor resections in about every fourth 2. Hall WA, Liu H, Martin AJ, Pozza CH, Maxwell RE, Truwit CL
patient. The extent of resection may only be a surrogate (2000) Safety, efficacy, and functionality of high field strength
parameter of clinical benefit for glioma patients. Yet, it interventional magnetic resonance imaging for neurosurgery. Neu
appears to be associated with prolonged survival, at least in rosurgery 46:632 641, discussion 641 642
3. Knauth M, Wirtz CR, Tronnier VM, Aras N, Kunze S, Sartor K
high-grade glioma patients [18 20]. Therefore, maximum safe (1999) Intraoperative MR imaging increases the extent of tumor
resection is the goal in brain tumor neurosurgery, and ioMRI resection in patients with high grade gliomas. AJNR Am J Neuror
emerges as an appropriate tool to achieve this goal. Effects on adiol 20:1642 1646
patient outcome in terms of survival, however, will yet have to 4. Nimsky C, Fujita A, Ganslandt O, Von Keller B, Fahlbusch R
(2004) Volumetric assessment of glioma removal by intraoperative
be demonstrated. high field magnetic resonance imaging. Neurosurgery 55:358 370,
discussion 370 371
5. Schulder M, Carmel PW (2003) Intraoperative magnetic resonance
imaging: impact on brain tumor surgery. Cancer Control 10:
Conclusion 115 124
6. Seifert V, Zimmermann M, Trantakis C, Vitzthum HE, Kuhnel K,
Raabe A, Bootz F, Schneider JP, Schmidt F, Dietrich J
We are conducting the first prospective randomized con- (1999) Open MRI guided neurosurgery. Acta Neurochir (Wien)
141:455 464
trolled trial involving the use of intraoperative image- 7. Senft C, Seifert V, Hermann E, Franz K, Gasser T (2008) Useful
guidance by means of ioMRI. The interim analysis of ness of intraoperative ultralow field magnetic resonance imaging
this study suggests that the use of ioMRI leads to lower in glioma surgery. Neurosurgery 63:257 266, discussion 266 257
Glioma Extent of Resection and Ultra-Low-Field ioMRI: Interim Analysis of a Prospective Randomized Trial 53
8. Schneider JP, Trantakis C, Rubach M, Schulz T, Dietrich J, shared resource magnetic resonance operating room after optimal
Winkler D, Renner C, Schober R, Geiger K, Brosteanu O, Zimmer image guided frameless stereotactic resection. Neurosurgery
C, Kahn T (2005) Intraoperative MRI to guide the resection of 48:731 742, discussion 742 744
primary supratentorial glioblastoma multiforme a quantitative 15. Oh DS, Black PM (2005) A low field intraoperative MRI system
radiological analysis. Neuroradiology 47:489 500 for glioma surgery: is it worthwhile? Neurosurg Clin N Am
9. Wirtz CR, Knauth M, Staubert A, Bonsanto MM, Sartor K, Kunze S, 16:135 141
Tronnier VM (2000) Clinical evaluation and follow up results 16. Schulder M (2009) Intracranial surgery with a compact, low field
for intraoperative magnetic resonance imaging in neurosurgery. strength magnetic resonance imager. Top Magn Reson Imaging
Neurosurgery 46:1112 1120, discussion 1120 1122 19:179 189
10. Hirschberg H, Samset E, Hol PK, Tillung T, Lote K (2005) Impact 17. Senft C, Franz K, Ulrich CT, Bink A, Szelenyi A, Gasser T,
of intraoperative MRI on the surgical results for high grade Seifert V (2010) Low field intraoperative MRI guided surgery of
gliomas. Minim Invasive Neurosurg 48:77 84 gliomas: A single center experience. Clin Neurol Neurosurg
11. Hadani M, Spiegelman R, Feldman Z, Berkenstadt H, Ram Z 112:237 243
(2001) Novel, compact, intraoperative magnetic resonance imaging 18. McGirt MJ, Chaichana KL, Gathinji M, Attenello FJ, Than K,
guided system for conventional neurosurgical operating rooms. Olivi A, Weingart JD, Brem H, Quinones Hinojosa A (2009)
Neurosurgery 48:799 807, discussion 807 809 Independent association of extent of resection with survival
12. Stummer W, Pichlmeier U, Meinel T, Wiestler OD, Zanella F, in patients with malignant brain astrocytoma. J Neurosurg
Reulen HJ (2006) Fluorescence guided surgery with 5 aminolevu 110:156 162
linic acid for resection of malignant glioma: a randomised con 19. Sanai N, Berger MS (2008) Glioma extent of resection and its
trolled multicentre phase III trial. Lancet Oncol 7:392 401 impact on patient outcome. Neurosurgery 62:753 764, discussion
13. Tuominen J, Yrjana SK, Katisko JP, Heikkila J, Koivukangas J 264 756
(2003) Intraoperative imaging in a comprehensive neuronavigation 20. Stummer W, Reulen HJ, Meinel T, Pichlmeier U, Schumacher W,
environment for minimally invasive brain tumour surgery. Acta Tonn JC, Rohde V, Oppel F, Turowski B, Woiciechowsky C,
Neurochir Suppl 85:115 120 Franz K, Pietsch T (2008) Extent of resection and survival in
14. Bohinski RJ, Kokkino AK, Warnick RE, Gaskill Shipley MF, glioblastoma multiforme: identification of and adjustment for
Kormos DW, Lukin RR, Tew JM Jr (2001) Glioma resection in a bias. Neurosurgery 62:564 576, discussion 564 576
Impact of a Low-Field Intraoperative MRI on the Surgical Results
for High-Grade Gliomas
Talat Kırış and Osman Arıca
Abstract In this study the authors retrospectively evaluated Introduction

the results of the operated intracranial high grade gliomas
using low field intraoperative MRI system Polestar N 20þ The prognosis of patients with high grade gliomas remains
Stealth Station (Medtronic, Co, USA) at German Hospital, poor despite advances in aggressive therapy with surgical
Istanbul. Between November 2006 and October 2008, 11 resection, radiotherapy and chemotherapy [1]. The goals
patients underwent microsurgical tumor resection with the of surgery in high grade gliomas are obtaining histopatho-
use of intraoperative MRI for WHO Grade III and IV glio- logical diagnosis, alleviating symptoms related increased
mas. There were six males and five females, mean age was intracranial pressure and compression of neural tissues,
53 (range 30 73), and mean follow-up duration was 19 increasing survival and decreasing the need for corticoster-
months (range 4 31). Ten total, one subtotal resection was oids. Growing evidence suggest that the extent of microsur-
achieved, whereas intraoperative MRI assesment demon- gical resection is associated with longer life expectancy in
strated five residual tumors. Histopathological examination high grade gliomas [2]. The availability of the intraoperative
revealed that eight tumors were Glioblastomas and three MRI has changed the principles of surgery for gliomas [3].
were anaplastic oligodendroglioma, anaplastic oligoastrocy- An intraoperative MRI scanner can acquire presurgical neu-
toma and anaplastic ependymoma respectively. No compli- roradiological examinations for planning, intraoperative
cations directly related to the intaoperative scanning were images for realtime neuronavigation and for comparing the
observed and there was no mortality, but one patient with an extent of tumor resection with preoperative tumor volume
insular tumor developed hemiparesis after the operation. [3 5]. Studies have shown improvement in the extent of
Mean hospital stay was 4.8 day. Ten patients received addi- tumor resection with the use of intraoperative MRI, but
tional radiotherapy and chemotherapy, one patient refused improved survival has not yet been proven [6 8]. In this
further therapy. Mean survival was 18.8 months for the study we report our results in neurosurgical resection of high
entire group and 15.6 months for glioblastoma patients. In grade gliomas in adults, with the help of a low field MRI
this small series of patients with high grade gliomas we scanner.
found that the use of intraoperative MRI helps complete
tumor removal and hence improves survival.
Keywords High grade glioma Intracranial tumors Intra-

operative MRI Microsurgery Patients and Methods
Within a period of 18 months between November 2007 and

June 2009, 13 craniotomies (for 11 patients) were performed
for removal of WHO grade III and IV Gliomas using the
low field intraoperative MRI scanner Polestar N-20, at the
T. Kırış (*) German Hospital, Istanbul. The group included 6 male and
Department of Neurosurgery, Istanbul School of Medicine, Istanbul 5 female patients with a mean age 53 (age range 30 73). One
University, 34093, Capa, Istanbul, Turkey
Department of Neurosurgery, German Hospital, Istanbul, Turkey
patient, operated earlier for a Grade II insular astrocytoma,
e mail: talatkrs@gmail.com had been operated two more times after the tumor was
O. Arıca recurred as Glioblastoma (the third operation was performed
Department of Neurosurgery, German Hospital, Istanbul, Turkey for a distant seeding metastasis).
56 T. Kırış and O. Arıca
Fig. 1 (continued)
Impact of a Low-Field Intraoperative MRI on the Surgical Results for High-Grade Gliomas 57
Fig. 1 (a) Preoperative T1 weighted postcontrast diagnostic images acquired with 1.5 T MRI scanner demonstrated a left deep seated frontal
tumor. (b) T1 weighted postcontrast images acquired with Polestar before surgery. (c) Neuronavigation using the images acquired with Polestar
before the beginning of surgery. (d) Intraoperative control MRI images for comparing the extend of resection. (e) Control MRI images 12 months
after the operation
For the surgical procedure, the patient was positioned at Results

the table with the non-magnetic headholder, designed for
this system. The scanner was placed under the operating In ten patients, total tumor resection was achieved. In one of
table. Patient reference frame and coil were attached to the the patients, brain stem infiltration prevented a total resec-
headholder and the patient’s head, respectively. The arms tion. Histopathological examination revealed that eight of
of the scanner were raised at a proper position. After the tumors operated were glioblastomas, one anaplastic oli-
the Starshield isolated the patient from the rest of the godendroglioma, one anaplastic oligoastrocytoma and one
room, postcontrast (double dose) T1 weighted images were anaplastic ependymoma. There was no mortality in this series.
obtained. This image set was used for planning the cranioto- The patient operated previously for low grade astrocytoma,
my. Tumors were removed using standard microsurgical developed hemiparesis after the second operation.
principles. Another intraoperative postcontrast T1 scanning Patients were positioned according to the surgical proce-
was obtained for evaluating the tumor removal. Whenever dure, either the prone, supine or lateral. The average time
rest tumor was detected, these images would be further used for patient positioning and image acquisition in this series
for real time neuronavigation. MRI scanning was repeated was 72 min (range 45 120 min). The average duration of
until complete resection was achieved (Fig. 1). surgery from incision to the last suture was 172 min. Intrao-
All patients except one (who refused any further therapy perative MRI demonstrated residual tumor in 5 of the 11
after microsurgery) received additional radiotherapy (a total cases. Imaging sessions averaged 2.4 per surgery. Mean
dose of 60 G). One patient recruited to a study at the Radia- hospital stay was 4.8 days.
tion Oncology Department of Istanbul University, received a Mean follow-up was 19 months (4 31months). Mean sur-
total dose of 66 G. Patients received radiotherapy, had also vival for the entire series was 18.8 months, mean survival for
received adjuvant chemotherapy; namely temozolamide. glioblastoma patients was 15.6 months. On the last follow-up,
58 T. Kırış and O. Arıca
four patients were disease free (2 Grade IV and 2 Grade III), Discussion
one patient with subtotal resection did not show any progres-
sion, three patients demonstrated recurrences and three For high grade gliomas, establishment of universally rele-
patients had died. One patient died of pneumonia 4 months vant prognostic criteria and treatment options remains a
after the operation, one patient who refused any further thera- great challenge. Tumor histopathology, patients’ age and
py died from recurrent tumor 7 months after the operation. functional status are the only reliable prognostic factors.
One patient died 28, 12 and 3 months after the first, second and Although Class I evidence is deficient in determining the
third operation, respectively. The second operation was per- efficacy of surgery in improving survival and delaying tumor
formed for recurrence and the third for a seeding metastasis. progression among patients with high grade gliomas, there is
No complications related to the use of intraoperative MRI growing evidence that more extensive surgical resection
had been observed. Table 1 summarizes the clinical features may be associated with more favorable life expectancy [2].
of the patients. Therefore some form of imaging modality like MRI, CT or
ultrasound, which increase the percentage of gross total
tumor resection should be of value.
During the past ten years, different intraoperative MRI
systems have been introduced into the neurosurgical
operating rooms, to allow realtime imaging during surgery
Table 1 Clinical features of the patients
[3, 5]. Intraoperative MRI systems with low- (0.15 0.5 T),
Sex
Female 5 high- (1.5 T) and ultrahigh- (3 T) field strength are available
Male 6 [3, 9 11]. High- and ultrahigh-field strength systems carry
Age potential of better imaging quality and opportunity of ad-
Mean 53 vanced imaging features, such as diffusion, spectroscopy and
Range 30 73 anjiography. However the size and cost of high- and ultra-
Location high-field strength magnets are pronounced disadvantages.
Teomporal 2
Polestar N-20 is a compact and mobile system with field
Insular 1
Frontal 2
strength of 0.15 T. The magnet is formed by two vertical,
Parietal 2 parallel disk shaped arms and the permanent magnet docks
Occipital 2 under a standard OR table. Because of its low magnetic field
Thalamic 1 strength, standard operating room and MRI-incompatible
Serebellum Brainstem 1 surgical instruments can be used. It provides more patient
Incision access since scanning and navigation are directly under the
Pterional 3
control of the neurosurgeon and thus eliminates the need
Horseshoe 1
Lineer 7
of simultaneous presence of a neuroradiologist or an MRI
Time for patient positioning and image acquisition technician.
Mean 45 120 min Schulder et al. reported their experience in cranial surgery
Range 72 min with the Polestar N-20 system [12]. Their report included
Imaging sessions 13 glioma cases. In their series, imaging sessions averaged
Mean 2.4 2.3 per surgery, which is almost identical with our experi-
Range 2 4 ence 2.4 average imaging per surgery. As they stated, we
Hospital stay
also found that the T1 weighted images (7 min/4 mm) are the
Mean 4.8 days
Histopathology most accurate for evaluating the extent of tumor resection as
Glioblastoma 8 well as for navigation. The extra time required for use of the
Anaplastic Oligodendroglioma 1 system, averaged 1.1 h in their series.
Anaplastic Oligoastrocytoma 1 Senft et al. reported their experience in glioma surgery
Anaplastic Epandymoma 1 (including low and high grade tumors) using the same sys-
Ki 67 index tem [8]. Patient positioning in their experience, took 33 min
Mean 31.15%
in average and additional time was used for image acquisi-
Range 9 80%
Follow up
tion. In our experience, patient positioningþimage acquisi-
Mean 19 tion lasted 72 min, which coincides with both reported series
Range 4 31 [11, 12]. Intraoperative scans after tumor resection revealed
Survival residual tumor in 47.6% of the patients with contrast en-
Mean for Total Series 18.8 months hanced tumors, in the series of Senft et al. In the present
Mean for Glioblastoma Patients 15.6 months series residual tumor was demonstrated in 45.4% (all high
Impact of a Low-Field Intraoperative MRI on the Surgical Results for High-Grade Gliomas 59
grade contrast enhanced tumors) and intentional gross total 3. Albayrak B, Samdani AF, Black PM (2004) Intraoperative mag
resection was achieved in 90.9% of patients. netic resonance imaging in neurosurgery. Acta Neurochir
146:543 557
Hirschberg et al. evaluated the impact of intraoperative 4. Iseki H, Nakamura R, Muragaki Y, Suzuki T, Chernov M, Hori T,
MRI on median survival for a series of patients harboring Takakura K (2008) Advanced computer aided intraoperative tech
high grade gliomas [13]. They compared this group to a nologies for information guided surgical management of gliomas:
matched cohort of patients operated in a conventional manner. Tokyo Women’s Medical University experience. Minim Invasive
Neurosurg 51(5):285 291
The intraoperative MRI scanner they used, was the Signa Sp/ 5. Oh DS, Black PM (2005) A low field intraoperative MRI system
I (General Electric Medical Systems, WI, USA), an open for glioma surgery: is it worthwhile? Neurosurg Clin N Am 16
vertical MRI scanner with 0.5 T field strength. The mean (1):135 141, Review
overall survival time in the study group was 14.5 months, 6. Knauth M, Wirtz CR, Tronnier VM, Aras N, Kunze S, Sartor K
(1999) Intraoperative MR imaging increases the extent of tumor
compared 12.1 months for the matched control group. With resection in patients with high grade gliomas. AJNR Am J Neuror
a similar system Schneider et al. evaluated the influence of adiol 20(9):1642 1646
intraoperative MRI on the extent of resection and the median 7. Schneider JP, Trantakis C, Rubach M, Schulz T, Dietrich J,
survival in patients with glioblastome multiforme [7]. The Winkler D, Renner C, Schober R, Geiger K, Brosteanu O, Zimmer C,
Kahn T (2005) Intraoperative MRI to guide the resection of prima
mean survival was 9.3 months, but when the patients were ry supratentorial glioblastoma multiforme a quantitative radio
analyzed separately for gross-total and subtotal resections, logical analysis. Neuroradiology 47(7):489 500
the mean survival was 17.9 and 7.9 months, respectively. In 8. Senft C, Seifert V, Hermann E, Franz K, Gasser T (2008) Usefull
our series, mean survival was 18.8 months for the whole ness of introperative ultra low field magnetic resonance imaging in
glioma surgery. Neurosurgery 63(4 suppl 2):257 266
group and 15.6 months for glioblastome multiforme patients. 9. Nimsky C, Ganslandt O, Fahlbusch R (2005) Comparing 0.2
The results of this small group of patients with high grade tesla with 1.5 tesla intraoperative magnetic resonance imaging
gliomas operated with the help of Polestar N-20, demon- analysis of setup, workflow and efficiency. Acad Radiol 12(9):
strated that a low field strength MRI scanner is a useful 1065 1079
10. Pamir MN, Ozduman K, Dinçer A, Yildiz E, Peker S, Ozek MM
surgical tool for evaluating the extent of tumor resection (2010) First intraoperative, shared resource, ultra high field 3 T
and for real time neuronavigation. magnetic resonance imaging system and its application in low
grade glioma resection. J Neurosurg 112(1):57 69
Conflict of interest statement We declare that we have no conflict 11. Sutherland GR, Kaibara T, Louw D, Hoult DI, Tomanek B, Saun
of interest. ders J (1999) A mobile high field magnetic resonance system for
neurosurgery. J Neurosurg 91(5):804 813
12. Schulder M, Salas S, Brimacombe M, Fine P, Catrambone J,
References Maniker AH, Carmel PW (2006) Cranial surgery with an expanded
compact intraoperative magnetic resonance imager. Technical
note. J Neurosurg 104(4):611 617
1. Oertel J, von Buttlar E, Schroeder HW, Gaab MR (2005) Prognosis 13. Hirschberg H, Samset E, Hol PK, Tillung T, Lote K (2005) Impact
of gliomas in the 1970s and today. Neurosurg Focus 18(4):E12 of intraoperative MRI on the surgical results for high grade glio
2. Sanai N, Berger M (2008) Glioma extent of resection and its mas. Minim Invasive Neurosurg 48(2):77 84
impact on patients outcome. Neurosurgery 62(4):753 766
Intraoperative MRI and Functional Mapping
Thomas Gasser, Andrea Szelenyi, Christian Senft, Yoshihiro Muragaki, I. Erol Sandalcioglu, Ulrich Sure,
Christopher Nimsky, and Volker Seifert
Abstract The integration of functional and anatomical data Introduction

into neuronavigation is an established standard of care in
many neurosurgical departments. Yet, this method has lim- The success of resective surgery of brain lesions in close
itations as in most cases the data are acquired prior to vicinity to eloquent cortex is closely linked to the preserva-
surgery. Due to brain-shift the accurate presentation of function or improvement of the initial neurofunctional status. For
tional as well as anatomical structures declines in the course that purpose, intraoperative electrophysiological monitoring
of surgery. (IOM) delivers information about the neurophysiological
In consequence, the acquisition of information during integrity of specific functional networks and pathways
surgery about the brain’s current functional state is of [1, 2]. Besides well-established and routinely employed
specific interest. The advancement of imaging technologies methods such as sensory and motor evoked potentials
(e.g. fMRI, MEG, Intraoperative Optical Intrinsic Signal (SEPs and MEPs), which test primarily the functional integ-
Imaging IOIS) and neurophysiological techniques and rity of the corticospinal tract, cortical mapping allows for
the advent of intraoperative MRI all had a major impact on testing of even more complex functional networks (e.g.
neurosurgery. speech, reading, cognition).
The combination of modalities such as neurophysiology Additionally, functional imaging technologies such as
and intraoperative MRI (ioMRI), as well as the acquisition functional MRI and magnetic source imaging (MEG) play
of functional MRI during surgery (ifMRI) are in the focus of an increasing role in the preoperative planning of the
this work. Especially the technical aspects and safety issues approach and the resection borders [3 7]. Postoperatively,
are elucidated. functional restitution and cortical reorganization may be
documented by fMRI, presenting further insight into issues
Keywords Brain mapping Functional MRI such as neuronal plasticity, which in turn is of growing
Intraoperative MRI Neurophysiology interest to neurosurgeons as well.
To date neuronavigation is the technical core structure for
T. Gasser (*) the integration of image based functional information into
Department of Neurosurgery, University of Duisburg Essen, Hufelandstr. the operative site. The usefulness of what may be called
55, 45147 Essen, Germany ‘‘functional neuronavigation’’ is well documented [1, 8 12]
Department of Neurosurgery, Goethe University, Schleusenweg 2 16, however, ‘‘functional neuronavigation’’ has technical limita-
60528 Frankfurt, Main, Germany
e mail: thomas.gasser@uk essen.de tions, primarily related to the fact, that the functional data
A. Szelenyi, C. Senft, and V. Seifert is acquired preoperatively. Loss of cerebrospinal fluid and
Department of Neurosurgery, Goethe University, Schleusenweg 2 16, tumour resection may result in a significant intraoperative
60528 Frankfurt, Main, Germany deformation of the brain resulting in an inaccuracy of
Y. Muragaki the neuronavigation, which in turn can be compensated
Department of Neurosurgery, Tokyo Women’s Medical University, for by intraoperative imaging and regular updates of the
Tokyo, Japan
navigational data sets [9 11].
I.E. Sandalcioglu, and U. Sure
Department of Neurosurgery, University of Duisburg Essen, Hufelandstr.
The study group has developed two approaches to address
55, 45147 Essen, Germany this specific issue in order to increase accuracy of intra-
C. Nimsky operative functional mapping. The technological pathways
Department of Neurosurgery, University of Marburg, Marburg, Germany being followed are:
62 T. Gasser et al.
Method 1: To perform functional MRI intraoperatively during and at the end of the surgical procedure. The block-
(intraoperative fMRI or ifMRI) employing a passive fMRI design stimulation paradigm alternated four rest and four
paradigm to visualize the sensorimotor cortex intraopera- activation periods of 33 s each. For functional imaging, slices
tively. parallel to the anterior posterior commissural plane were
This method has been implemented in a 1.5 T, as well as acquired as T2*-weighted Echo Planar Imaging (EPI)
in a 0.3 T environment [13]. sequences (TE 60 ms, TR 3,300 ms, flip-angle 90 , slice
Method 2: To routinely update intraoperative anatomical thickness 3 mm, FOV 220 mm, matrix 6464) with data
neuronavigation and to combine these data with conventional acquisition covering the whole cerebrum.
IOM. This method has been applied in a 0.15 T setup [12]. Employing Statistical Parametric Mapping (SPM, Well-
come Department of Imaging Neuroscience, London, Ver-
sion 99), the functional time series were analyzed
statistically. The data were motion corrected by rigid body
Materials and Methods spatial translation. After data smoothing, a design matrix
according to the variables of the above mentioned boxcar
stimulus function was created. The predictor vectors were
Method 1: Intraoperative Functional MRI convolved with a hemodynamic response function. The data
were assessed statistically specifying P as 0.05 (n¼1) for
In the presented study, four patients (mean age: 39 years) corrected and 0.001 (n¼3) for uncorrected comparisons.
with centrally located lesions (one glioblastoma WHO For anatomical correlation the functional data with the
IV, one astrocytmoma WHO II, one bronchial carcinoma realigned echo planar images were co-registered with a
metastasis, one cavernoma) were evaluated under total intra- 3-D T1-weighted high-resolution data set.
venous anaesthesia (TIVA) with propofol 2% (2,6-diisopro- A comparable procedure to the one described above
pylphenol; average dosage: 5.8 mgkg 1 h 1) and fentanyl has been applied by the study group in a different setting,
(1-phenethyl-4-(phenylpropionylamino)piperidine; average namely in a 0.3 T ioMRI environment (Hitachi AIRIS II,
dosage: 0.23 mgkg 1 h 1) at different stages of surgery. In Hitachi Medical Cooperation, Kashiwa, Japan) which is
order to acquire functional MRI, several technical prerequi- installed in the operating suite of the Department of Neuro-
sites have to be met. Firstly, the installation of an intrao- surgery at Tokyo Women’s Medical University, Tokyo,
perative MR unit (ioMRI) is required. The neurosurgical Japan. Minor modifications to the scanner software (Version
operating suite in the Department of Neurosurgery, Univer- 5.0 M) and an adaptation of the scanning parameters (echo
sity Erlangen-Nuremberg, is equipped with such a 1.5-T planar sequence GE-EPI, TR 1,200 ms and TE 41.8 ms, flip
scanner (Magnetom Sonata Maestro Class, Siemens Medical angle 80 ) were necessary to implement ifMRI.
Solutions, Erlangen, Germany). Secondly, a passive func-
tional MR-paradigm, eliciting an activation of the sensori-
motor cortex under general anaesthesia has to be employed.
The study group has described this method’s setup, safety Method 2: Combination of ioMRI and IOM
and scope of application in detail previously [8, 13, 14]. In
short, we applied a passive paradigm based on electrical The combination of intraoperative MRI (ioMRI), neurona-
stimulation of the median and tibial nerves [14]. For electri- vigation, intraoperative neurophysiological localization (e.g.
cal stimulation an electromagnetically shielded coaxial lead cortical mapping, phase reversal) and continuous functional
(length: 8 m; resistance: 50 O) was designed. Shielding was monitoring promises to the neurosurgeon up-to-date func-
achieved by connecting the conductor’s shielding-mesh to tional and anatomical information. However, the interaction
the MRI-cage. Modified bipolar surface electrodes (Nicolet, between the magnetic field (either static or alternating) and
Madison, WI, USA; Part Number: 019-401500) were utilized the electrodes, which are inserted in the scalp, may induce
for both stimulated locations, applying 3-Hz square wave heat or an electric current. The maximum interaction within
electrical pulses of 100 ms duration with an intensity of this specific ioMRI environment (PoleStar 0.15-T, Medtro-
3 mA above motor threshold. The impulse generator (Nicolet- nic Surgical Navigation Technologies, Louisville, CO,
Viking IV P, Madison, WI, USA) was located outside the USA) and the maximum electric current being delivered by
operating suite and the conductor was threaded through a the IOM system (ISIS intraoperative monitoring system,
waveguide array into the operating theatre. After anaesthe- Inomed Co., Teningen, Germany) were first of all estimated
siological induction and patient positioning, the stimula- in a precursor study evaluating effects such as induction of
tion electrodes were attached and the motor threshold was heat or directional forces in a phantom. Additionally, the
defined. After an initial anatomical and functional MR scan, effects on image quality were evaluated. The MRI data sets
surgery commenced and two further data sets were acquired included the following axial sequences: T1-weighted
Intraoperative MRI and Functional Mapping 63
Fig. 1 Phantom prepared for standardized evaluation of artifacts originating from different electrodes. Left frame: comparison of the image
artefacts, with the corkscrew electrodes producing a considerable signal wipe out (from left to right: no electrode; platinum iridium (PtIr)
electrode; corkscrew electrode; steel electrodes. Right frame: actual setup with phantom placed in between the MR unit’s aperture (Polestar N20)
with electrode attached to the phantom’s calibration sphere (white arrow)
(TR 40.00 ms, TE 3.00 ms, with and without gadolinium- ioMRI. In 15 patients we additionally mapped the cortex
DTPA) and T2-weighted (TR 3,000.00 ms TE 112 ms). with direct cortical stimulation (DCS). For IOM Pt/Ir elec-
After establishing safety criteria, the clinical feasibility was trodes were utilized (Fig. 2). ioMRI was performed at the
evaluated in a subsequent study. start of surgery and at the presumed end of resection employ-
ing the following axial sequences: T1-weighted (TR
40.00 ms, TE 3.00 ms, with and without gadolinium-DTPA),
T2-weighted (TR 3,000.00 ms, TE 112 ms).
Phantom Study
During the scanning procedure the ION System was
switched off and the head box was detached from the
Standard CE-certified electrodes (stainless steel subdermal
amplifier.
electroencephalography [EEG] needle electrodes, corkscrew
design electrodes, and platinum/iridium [Pt/Ir]-subdermal
EEG needle electrodes; Viasys Healthcare, Madison, WI;
0.4-mm diameter and 1.5-m lead length) were evaluated,
by attaching the electrodes to a copper sulfate-filled phantom
Results
(PoleStar N20 specific equipment, Medtronic Surgical Nav-
igation Technologies, Louisville, CO, USA) (Fig. 1). While Results of Method 1: Intraoperative
imaging was performed with the PoleStar N20 (Medtronic Functional MRI
Surgical Navigation Technologies, Louisville, COS, USA;
scanning parameters, see above), induction of movement
In all patients examined with the 1.5 T ioMRI, the somato-
and heat was observed and measured. Image quality was
sensory cortex could be identified by intraoperative fMRI.
analyzed for artifacts by comparing images with and without
The signal characteristics varied in the course of surgery
electrodes.
with changing BOLD signal intensities and co-activations
of the assumed secondary sensory cortex (S2). Remarkably,
an inversal of the BOLD response in nearly half of the
Clinical Study measurements was observed. Analyzing the functional data
sets by introducing a contrast of 1 suggested an increase of
In 29 patients (median age: 40 years) with supratentorial the deoxyhemoglobin concentration in the elicited somato-
lesions (11 glioblastoma WHO IV, 15 astrocytoma WHO sensory cortex. Microscope-based neuronavigation allowed
II-III, 1 cavernoma, 1 tuberculoma, oligodendroglioma) a the intraoperative correlation of anatomical and functional
neurosurgical procedure was performed with IOM and data with the exposed structures in the surgical field.
64 T. Gasser et al.
were no adverse effects related to the setup and the quality

of the monitoring was not affected either.
Discussion
The integration of preoperatively acquired functional imag-

ing and intraoperative neurophysiology employing neurona-
vigation has been implemented in several centres. However,
only a few reports exist about intraoperative updates of
functional information in combination with ioMRI [12].
Two different technical setups, which provide such a func-
tional update, are in focus of the presented data.
Fig. 2 Placement of the Platinum Iridium (PtIr) electrodes for IOM Established methods of functional mapping and monitor-
according to the international 10 20 EEG system in a patient with a ing, such as direct cortical stimulation, phase reversal or
recurrent astrocytoma. Note the location of electrodes within the
HF head coil (Medtronic Surgical Navigation, Louisville, CO, USA). evoked potentials, have still to be regarded as the gold
To avoid heating and dislocation of the electrodes induced by the static standard of brain mapping. The combination with ioMRI
and alternating magnetic field, all cables were braided, and looping of could however facilitate the interpretation of neurophysio-
the electrode cables was avoided. The preamplifier was placed at a logical information. Thus efforts were made to combine the
distance of 1 m from the patient’s head at the sidebar of the operating
table placed (not visible) technical components as described under method 2. The
advantage of this method is, that it offers the high temporal
The average registration accuracy was 1.45 mm (standard resolution and the constant flow of information originating
error: 0.67 mm). Significant susceptibility artifacts were from IOM and electrophysiological mapping. In our setup,
avoided by filling the resection cavity with saline. Applying we have proven safety only for a 0.15 T environment, with
SPM’s statistical routines, the fitted response function (ad- its obviously reduced spatial resolution and the smaller field-
justed data versus time/scans) documented in all cases a high of-view when compared to a 1.5 T MR unit. With careful
correlation between the temporal course of the paradigm and selection of the electrode types, EEG and evoked potential
the cortical activation. recording, even in high-field MRI systems up to 7-T, has
In a preliminary evaluation we could demonstrate that been described as safe [15].
similar results were obtained in patients in whom intraopera- The primary assets of ioMRI are the compensation of
tive fMRI was performed employing a 0.3 T scanner. brain-shift and intraoperative resection control. Extended
capabilities of ioMRI, namely functional MRI (fMRI) and
diffusion tensor imaging (DTI), offer additional information
to the neurosurgeon. To date, functional navigation relies on
Results of Method 2: Combination of ioMRI preoperatively acquired data. However, as described under
and IOM method 1 the intraoperative acquisition and update of func-
tional MRI is safe and feasible. In comparison to method 2,
Phantom Study which propagates the combination of 0.15 T ioMRI and
IOM, the functional information is updated only at a few
Primarily steel electrodes caused electrode-related artifacts. times during surgery. This may be a disadvantage, yet the
The most common artifacts were signal deletions and local high anatomical resolution and the combination of up-to-
field distortions. More importantly, this study demonstrated date intraoperative fMRI and diffusion tensor imaging (DTI)
the inertness of Pt/Ir electrodes to MRI (Fig. 1); consequently, provide a technical solution. By co-registration of ifRMI and
they were used for the patient study. Temperature changes DTI, the entire intracranial corticospinal tract can be visua-
were not observed in any of the utilized electrodes. lized [16]. The application of ifMRI activation areas as seed
points for fiber tracking algorithms in DTI data in order to
delineate major white matter tracts intraoperatively can pro-
Clinical Study vide more exact data, reducing the risk for postoperative
neurological deficits. Especially in surgery of larger, centrally
In all collected ioMRI data sets, irrespective of the sequence, located lesions with an increased shift of the adjacent gyri,
no image artifacts or changes in SNR (signal-to-noise-ratio) the update of functional information may help to trace the
originating from the electrodes could be detected. There sensorimotor cortex.
Intraoperative MRI and Functional Mapping 65
Conclusion functional magnetic resonance imaging data in a neuronavigational

system. Neurosurgery 49:1145 1157
7. Towle VL, Khorasani L, Uftring S, Pelizzari C, Erickson RK,
Neurophysiological monitoring with evoked potentials and Spire JP, Hoffmann K, Chu D, Scherg M (2003) Noninvasive
DCS can be performed safely and with good quality within a identification of human central sulcus: a comparison of gyral
low-field open MRI system. Alternatively, intraoperative morphology, functional MRI, dipole localization, and direct corti
fMRI, employing a 1.5 T MR unit and an appropriate para- cal mapping. Neuroimage 19:684 697
8. Gasser T, Sandalcioglu IE, Schoch B, Gizewski ER, Forsting M,
digm, represents an equally safe and technically feasible Stolke D, Wiedemayer H (2005) Functional MRI in anaesthetized
method for near-real-time identification of eloquent brain patients. A relevant step towards real time intraoperative functional
areas despite brain shift during neurosurgical procedures. neuroimaging. Neurosurgery 57:94 99
9. Nabavi A, Black PM, Gering DT, Westin CF, Mehta V,
Disclosure Thomas Gasser, M.D., Ph.D., serves as clinical consul Pergolizzi RS Jr, Ferrant M, Warfield SK, Hata N, Schwartz RB,
tant to Medtronic, but there has been no direct financial support of this Wells WM 3rd, Kikinis R, Jolesz FA (2001) Serial intraopera
study by the company. tive magnetic resonance imaging of brain shift. Neurosurgery 48:
787 798
10. Nimsky C, Ganslandt O, Cerny S, Hastreiter P, Greiner G,
References Fahlbusch R (2000) Quantification of, visualization of, and com
pensation for brain shift using intraoperative magnetic resonance
imaging. Neurosurgery 47:1070 1080
1. Berger MS, Rostomily RC (1997) Low grade gliomas: functional 11. Nimsky C, Ganslandt O, Hastreiter P, Fahlbusch R (2001)
mapping resection strategies, extent of resection, and outcome. J Intraoperative compensation for brain shift. Surg Neurol 56:
Neurooncol 34:85 101 357 365
2. Neuloh G, Pechstein U, Cedzich C, Schramm J (2004) Motor 12. Szelenyi A, Gasser T, Seifert V (2008) Intraoperative neurophysi
evoked potential monitoring with supratentorial surgery. Neuro ological monitoring in an open low field magnetic resonance
surgery 54:1061 1072 imaging system: clinical experience and technical considerations.
3. Achten E, Jackson GD, Cameron JA, Abbott DF, Stella DL, Neurosurgery 63:268 275
Fabinyi GC (1999) Presurgical evaluation of the motor hand area 13. Gasser T, Ganslandt O, Sandalcioglu IE, Stolke D, Fahlbusch R,
with functional MR imaging in patients with tumors and dysplastic Nimksky C (2005) Intraoperative functional MRI: implementation
lesions. Radiology 210:529 538 and preliminary experience. Neuroimage 26:685 693
4. Hill DL, Smith AD, Simmons A, Maurer CR Jr, Cox TC, Elwes R, 14. Gasser T, Sandalcioglu IE, Wiedemayer H, Hans V, Gizewski ER,
Brammer M, Hawkes DJ, Polkey CE (2000) Sources of error in Forsting M, Stolke D (2004) A novel passive functional MRI
comparing functional magnetic resonance imaging and invasive paradigm for preoperative identification of the somatosensory
electrophysiological recordings. J Neurosurg 93:214 223 cortex. Neurosurg Rev 27:106 112
5. Kober H, Nimsky C, Moller M, Hastreiter P, Fahlbusch R, 15. Vasios CE, Angelone LM, Purdon PL, Ahveninen J, Belliveau JW,
Ganslandt O (2001) Correlation of sensorimotor activation with Bonmassar G (2006) EEG/(f)MRI measurements at 7 Tesla using a
functional magnetic resonance imaging and magnetoencephalogra new EEG cap (‘‘InkCap’’). Neuroimage 33:1082 1092
phy in presurgical functional imaging: a spatial analysis. Neuro 16. Nimsky C, Ganslandt O, Hastreiter P, Wang R, Benner T,
image 14:1214 1228 Sorensen AG, Fahlbusch R (2005) Intraoperative diffusion tensor
6. Roux FE, Ibarrola D, Tremoulet M, Lazorthes Y, Henry P, Sol JC, MR imaging: shifting of white matter tracts during neurosurgical
Berry I (2001) Methodological and technical issues for integrating procedures initial experience. Radiology 234:218 225
Information-Guided Surgical Management of Gliomas
Using Low-Field-Strength Intraoperative MRI
Yoshihiro Muragaki, Hiroshi Iseki, Takashi Maruyama, Masahiko Tanaka, Chie Shinohara,
Takashi Suzuki, Kitaro Yoshimitsu, Soko Ikuta, Motohiro Hayashi, Mikhail Chernov,
Tomokatsu Hori, Yoshikazu Okada, and Kintomo Takakura
Abstract (263 cases; 45.8%) were followed more than 2 years after
Background: Contemporary technological developments surgery.
revolutionized management of brain tumors. The experience Findings: Maximal possible tumor resection, defined as
with information-guided surgery of gliomas, based on the radiologically complete tumor removal or subtotal removal
integration of the various intraoperative anatomical, func- leaving the residual neoplasm within the vital functionally-
tional, and histological data, is reported. important brain areas, was attained in 569 cases (99.1%).
Methods: From 2000 to 2009, 574 surgeries for intracra- The median resection rate constituted 95%, 95%, and 98%,
nial gliomas were performed in our clinic with the use of for WHO grade II, III, and IV gliomas, respectively. Actuar-
intraoperative MRI (ioMRI) with magnetic field strength of ial 5-year survival was significantly worse in WHO grade IV
0.3 T, updated neuronavigation, neurochemical navigation gliomas (19%), but did not differ significantly between
with 5-aminolevulinic acid, serial intraoperative histopatho- WHO grade III and II tumors (69% vs. 87%).
logical investigations of the resected tissue, and comprehen- Conclusions: Information-guided management of glio-
sive neurophysiological monitoring. Nearly half of patients mas using low-field-strength ioMRI provides a good oppor-
tunity for maximal possible tumor resection, and may result
in survival advantage, particularly in patients with WHO
Y. Muragaki (*), T. Maruyama, and M. Hayashi grade III neoplasms.
Faculty of Advanced Techno Surgery, Institute of Advanced
Biomedical Engineering and Science, Graduate School of Medicine,
Tokyo Women’s Medical University, 8 1 Kawada cho, Shinjuku ku, Keyword Brain tumor resection Glioma Intraoperative
Tokyo, 162 8666, Japan MRI Surgery Updated intraoperative neuronavigation
Department of Neurosurgery, Neurological Institute, Tokyo Women’s
Medical University, 8 1 Kawada cho, Shinjuku ku, Tokyo, 162 8666,
Japan
e mail: ymuragaki@abmes.twmu.ac.jp
Introduction
H. Iseki, M. Chernov, and K. Takakura
Biomedical Engineering and Science, Graduate School of Medicine, Contemporary technological developments revolutionized
Tokyo Women’s Medical University, 8 1 Kawada cho, Shinjuku ku, surgical management of intraaxial brain tumors. Introduc-
Tokyo, 162 8666, Japan
Department of Neurosurgery, Neurological Institute, Tokyo Women’s tion of the intraoperative MRI (ioMRI) and related updated
Medical University, 8 1 Kawada cho, Shinjuku ku, Tokyo, 162 8666, neuronavigation permitted for neurosurgeons to perform
Japan resection of the tumor under precise anatomical guidance.
International Research and Educational Institute for Integrated Medical Moreover, at present it is possible to obtain intraoperatively
Sciences (IREIIMS), Tokyo Women’s Medical University, 8 1
Kawada cho, Shinjuku ku, Tokyo, 162 8666, Japan not only structural, but functional and metabolic images.
M. Tanaka, C. Shinohara, T. Hori, and Y. Okada,
Comprehensive neurophysiological monitoring and intrao-
Department of Neurosurgery Neurological Institute, Tokyo Women’s perative brain mapping by direct cortical and subcortical
Medical University, 8 1 Kawada cho, Shinjuku ku, Tokyo, 162 8666, electrical stimulation, particularly performed during ‘‘awake
Japan craniotomy’’, allows precise localization of the cerebral func-
T. Suzuki, K. Yoshimitsu, and S. Ikuta tions and preservation of the functionally-important brain
Biomedical Engineering and Science, Graduate School of Medicine,
structures during removal of the tumor. Neurochemical navi-
Tokyo Women’s Medical University, 8 1 Kawada cho, Shinjuku ku, gation with 5-aminolevulinic acid (5-ALA) permits direct
Tokyo, 162 8666, Japan visualization of the residual neoplasm under the operating
68 Y. Muragaki et al.
microscope and its differentiation from the peritumoral brain, surgical point of view it allows use of any required approach
whereas techniques of intraoperative histopathological including retrosigmoid and transtentorial. Moreover, use of
diagnosis allow fast direct investigation of the pathological the device provides an opportunity to perform intraopera-
tissue for identification of the neoplastic cells. Incorporation tively not only volumetric, but diffusion-weighted imaging
of these adjuncts into surgical management of gliomas pro- (DWI), MR angiography (MRA), and functional investiga-
vides for the surgeon an opportunity to perform aggressive tions, with sufficient quality of images comparable to
resection of the tumor with minimal risk of postoperative those one obtained on scanners with higher magnetic field
neurological morbidity. The present report highlights our strength.
experience with information-guided surgical management For facilitation of the tumor removal and detection of the
of gliomas using low magnetic field strength ioMRI with neoplastic remnants we use previously developed navigator
an emphasis on tumor resection rate and outcome. for photon radiosurgery system (PRS navigator; Toshiba,
Tokyo, Japan), which allows fast and easy updating of the
information obtained with ioMRI. It is based on a conven-
tional infrared location-identification device, which shows
Materials and Methods the location of the suction tip and position of the suction tube
in 3 sectional planes. The mislocalization errors of the de-
From 2000 to 2009, 734 neurosurgical procedures with the vice constitute 0.8 mm in average, 1.5 mm at maximum, and
use of ioMRI were performed in the intelligent operating 0.5 mm at minimum, and typically do not exceed 1 mm [1].
theater of the Tokyo Women’s Medical University. The vast The system permits co-registration, fusion and three-dimen-
majority of these surgeries (574 cases; 78.2%) were directed sional reconstruction of the various images, and provides
on biopsy or resection of gliomas using the concept of their easy-to-understand information.
information-guided surgical management. Nearly half of If the tumor is located near or in the eloquent brain area,
patients (263 cases; 45.8%) were followed more than cortical mapping, neurophysiological monitoring, and/or
2 years after surgery. stimulation of the cranial nerves are performed as appropri-
ate before resection of the neoplasm for identification of
the motor area, speech area, cranial nerves and their nuclei.
‘‘Awake craniotomy’’ was performed in 152 cases of the present
Intelligent Operating Theater series. Motor evoked potentials were investigated in 437
cases, whereas somatosensory evoked potentials were moni-
Detailed information on the internal organization of the tored routinely during surgery. Neurochemical navigation
intelligent operating theater of the Tokyo Women’s Medical with 5-ALA and intraoperative histopathological investiga-
University has been provided previously elsewhere [1, 2]. It tion were done routinely, as appropriate.
equipped with an open ioMRI scanner (AIRIS IITM, Hitachi
Medical Co., Chiba, Japan) with magnetic field strength of
0.3 T, which has a hamburger-like shape with a 43 cm gantry
Concept of the Information-Guided Surgical
gap and a permanent magnet producing vertical magnetic
field with resonance frequency of 12.7 MHz. Low magnetic Management of Gliomas
field strength creates narrow 5-gauss line, which extends 2 m
from both sides, 2.2 m in front, 1.8 m backwards, and 2.5 m Our concept of the information-guided surgical management
upwards. It permits for the surgeon to use some conventional of gliomas is based on the integration of various intraopera-
surgical devices and instruments in the working space out- tive anatomical, functional, and histological data with a
side 5-gauss line. It should be specially marked, that this purpose to attain maximal surgical resection of the tumor
ioMRI scanner does not require a cooling system, which with minimal risk of postoperative neurological morbidity
significantly reduce its operating costs by approximately (Fig. 1). In our practice ioMRI investigations are usually
10,000 Japanese yen (around 100 US $) per month. Origi- performed when approach to the tumor is attained and then
nally developed radiofrequency receiver coil integrated with when the lesion is removed [2]. It permits assessment of
Sugita head-holder (Head-holder coil; Mizuho Ltd., Tokyo, completeness of the tumor resection and identification of the
Japan) significantly improved the quality of intraoperative residual neoplasm or possible adverse effects, such as hae-
images. It provides an opportunity to perform MRI investi- morrhage. If residual tumour is identified and considered
gations with minimal distortion artifacts and maximum suitable for additional resection the newly obtained ioMRI
structure contrasting in any plane irrespectively to orienta- data are transferred to the neuronavigation device with
tion of the object, which allows fixation of the patient head in subsequent resection of the neoplasm using this updated
the most desirable position for tumor removal. From the information.
Information-Guided Surgical Management of Gliomas Using Low-Field-Strength Intraoperative MRI 69
Fig. 1 Main principles of

information guided surgery for
glioma. To maximize resection
rate and minimize neurological
morbidity various types of
intraoperative information,
namely anatomical, functional,
and histological, are analyzed. All
data are integrated with updated
neuronavigation. In surgical
decision making the first priority
is given to functional information
provided by intraoperative brain
mapping and comprehensive
neurophysiological monitoring.
ioMRI intraoperative MRI, MEP
motor evoked potentials, SEP
somatosensory evoked potentials
Whereas such anatomical data are used for removal of (Fig. 2). Importantly, our low-field-strength ioMRI showed
the bulk of the tumor, the resection of the residual neoplasm high sensitivity for detection of the residual glioma, which
is based not only on the anatomical images, but on the was confirmed by postoperative high-field-strength MRI
results of the neurochemical navigation with 5-ALA and investigations. In no one case of the present series unexpect-
histopathological investigation of the walls of the surgical ed residual tumor was disclosed.
cavity. Actuarial 5-year survival rate in patients who were fol-
It should be marked, that neither anatomical data lowed more than 2 years after surgery was significantly
obtained with ioMRI, nor histopathological information on worse in WHO grade IV gliomas (19%; P<0.0001), but
presence of the residual neoplasm are sufficient for guidance did not differ significantly between WHO grade III and II
of the tumor resection, especially if the lesion is located near tumors (69% vs. 87%; P¼0.0942).
or within functionally-important cerebral structures. In our
opinion in any occasion the first priority in surgical decision-
making should be given to functional information provided Illustrative Case
by the intraoperative brain mapping and comprehensive
neurophysiological monitoring.
An 18-year-old man underwent information-guided surgery
for left insular glioblastoma multiforme located adjacent to
the compressed pyramidal tract medially and arcuate fascic-
ulus associated with speech function, superiorly (Fig. 3).
Results After craniotomy and dissecting of the Sylvian fissure
baseline iMR images were obtained. Tumor removal was
ioMRI investigations provided informative images in 572 attained under awake condition of the patient. Incorporation
out of 574 cases (99.7%). of the DWI into neuronavigation device permitted for the
Maximal possible tumor resection was attained in 569 surgeon to identify clearly the white matter bundle contain-
patients (99.1%). It included cases of radiologically complete ing pyramidal tract and to perform anatomically controlled
tumor removal as well as subtotal removal leaving the residu- dissection of the neoplasm from this functionally important
al neoplasm within the vital functionally-important brain structure. Transitory motor weakness and aphasia were
areas detected with neurophysiological monitoring and/or observed at the time of surgical manipulations in the vicinity
brain mapping. In 3 cases aggressive tumor resection was to pyramical tract and superior longitudinal tract, respectively.
not completed as planned due to lost of cooperation with the Total en bloc removal of the contrast-enhanced tumor was
patient operated on in awake condition (2.0% of such cases). attained and control ioMRI did not disclose residual neo-
The resection rate did not depend on the histopathological plasm. Correspondingly, histopathological investigation of
tumor grade and constituted, in median, 95%, 95%, and the wall of the resection cavity did not disclose neoplastic
98%, for WHO grade II, III, and IV gliomas, respectively cells’ clusters. The postoperative period was uncomplicated.
Fig. 2 Results of information guided surgery for glioma using 0.3 T intraoperative MRI. Images before (upper row) and after (lower row) tumor
removal are presented: (a) 100% resection of glioblastoma multiforme of the corpus callosum and both frontal lobes; (b) 85% resection of
glioblastoma multiforme of the left thalamus; (c) 98% resection of the low grade astrocytoma of the left basal ganglia and insula; (d) 98%
resection of the anaplastic astrocytoma of the left parietal lobe with residual tumor left in the functionally important brain area (arrowhead)
Transitory intraoperative motor weakness and aphasia re- 4]. The advantages of such devices include high image quali-
covered completely and no neurological morbidity was ob- ty, possibility to attain diffusion tensor and spectroscopic
served after surgery. images, and short scanning time. However, increase of mag-
netic field strength is associated with greater possibility of
image distortion artifacts. Additionally, the maintenance costs
of such iMR scanners is high. In the same time latest technical
Discussion achievements have permitted improvement of image quality
of low-field-strength ioMRI [1, 5 7]. As it is shown herein,
Introduction of ioMRI and updated neuronavigation into our open ioMRI scanner with magnetic field strength of just
routine neurosurgical practice provided an opportunity to 0.3 T allows to attain not only standard T1-weighted and T2-
perform resection of gliomas under precise anatomical guid- weighted images of sufficient quality and resolution, but
ance. In our initial report use of ioMRI resulted in 93% permits one to perform functional, DWI and MRA investiga-
average resection rate of gliomas, and 46% of total tumor tions. Further possibilities of fusion of preoperative high-
removals [2]. Growing experience resulted in even better field-strength and intraoperative low-field-strength MR
results presented herein. Maximal possible tumor resection, images using deformable model within advanced computer
defined as radiologically complete tumor removal or subto- software, can result in combining of the advantages of both
tal removal leaving the residual neoplasm within the vital techniques and may significantly diminish the need for ioMRI
functionally-important brain areas, was attained in 99.1% of of high magnetic field strength.
cases. While such aggressive tumor resection is frequently It should be specially underlined, that in cases of intra-
associated with increased rate of temporary postoperative axial brain neoplasms anatomical data alone, even if
neurological deterioration, the permanent neurological mor- obtained with ioMRI, are not sufficient for guidance of
bidity was noted just in 14% of cases [2]. Additionally, it the resection. Precise histopathological information and,
seems that use of ioMRI reduces the risk of postoperative especially, use of comprehensive neurophysiological mon-
hemorrhagic complications, and does not increase the risk of itoring with cortical and subcortical brain mapping are
infection despite prolonged operative time [2]. absolutely necessary for attainment of the optimal results
During the last decade there is a trend for introduction of [2, 8]. Intraoperative integration of anatomical, histopatholo-
ioMRI with high magnetic field strength of 1.5 T and 3 T [3, gical and neurophysiological data constitutes the basis of our
Information-Guided Surgical Management of Gliomas Using Low-Field-Strength Intraoperative MRI 71
Fig. 3 Information guided

surgery for left insular
glioblastoma multiforme in 18
year old man. Contrast enhanced
T1 weighted intraoperative MR
image obtained after craniotomy
and before tumor resection
showed the tumor located in the
left posterior insula (a).
Incorporation of the
intraoperative diffusion weighted
image into neuronavigation
system permitted for the surgeon
to identify white matter bundle
containing pyramidal tract (b,
arrow), and brain areas in which
surgical manipulations resulted in
transitory motor weakness (c) and
aphasia (d). At the end of tumor
removal no definite residual
neoplasm could be seen on
structural intraoperative MRI (e)
and, correspondingly,
histopathological investigation of
the wall of the resection cavity
did not disclose neoplastic cells’
clusters (f)
concept of ‘‘information-guided surgery’’ with a goal of median overall survival after complete removal of the
maximal possible tumor resection and minimal risk of postcontrast-enhanced lesion (17 months) was significantly
operative neurological deterioration. longer compared to cases with its incomplete removal (12
Meanwhile, the main question is still remains unanswered: months). In concordance, in the report of van den Bent et al.
whether total removal of the glial neoplasm, particularly high- [10] on EORTC 26951 randomized trial of combined che-
grade one, is translated into longer survival of the patient? motherapy for anaplastic gliomas, the overall survival was
Published data on this important issue are unequivocal and better after complete tumor removal compared to partial
result in a great controversy providing few reasonable ones or to biopsy. The similar results were marked in the
arguments both for opponents and advocates of aggressive Brain Tumor Registry of Japan [11]. Analysis of 6,400 cases
tumor resection. It seems, however, that the study of Stummer of WHO grade III and IV gliomas showed, that more than
et al. [9] provides a high level of evidence in favor of total 90% tumor removal is associated with survival advantage,
glioma removal. The authors adjusted biases of age and while such resection rate was attained in 6 10% of cases
eloquent area location in the dataset of randomized study only.
on use of neurochemical navigation with 5-ALA during Our current retrospective analysis could not be reliable
resection of glioblastoma multiforme, and found that for evaluation of the association between resection rate of
glioma and patients’ survival after surgery. Anyway, in the Tokyo Women’s Medical University experience. Minim Invasive
present series actuarial 5-year survival rate for patients with Neurosurg 51:285 291
2. Muragaki Y, Iseki H, Maruyama T, Kawamata T, Yamane F,
WHO grade II, III, and IV gliomas constituted 87%, 69%, Nakamura R, Kubo O, Takakura K, Hori T (2006) Usefulness of
and 19%, respectively. For comparison, according to the last intraoperative magnetic resonance imaging for glioma surgery.
edition of the Brain Tumor Registry of Japan [11] the same Acta Neurochir Suppl 98:67 75
rates in general neurosurgical practice constitutes 75%, 40%, 3. Nimsky C, Ganslandt O, Von Keller B, Romstock J, Fahlbusch R
(2004) Intraoperative high field strength MR imaging: implemen
and 7%, respectively. Moreover, the finding of the similar tation and experience in 200 patients. Radiology 233:67 78
long-term prognosis in our cases of WHO grade II and III 4. Pamir MN, Ozduman K, Dincer A, Yildiz E, Peker S, Ozek MM
gliomas seems intriguing. It can reflect the significance of (2010) First intraoperative, shared resource, ultrahigh field 3 Tesla
survival advantage associated with aggressive resection of magnetic resonance imaging system and its application in low
grade glioma resection. J Neurosurg 112:57 69
WHO grade III gliomas [12] and should be definitely inves- 5. Hadani M, Spiegelman R, Feldman Z, Berkenstadt H, Ram Z
tigated further. (2001) Novel, compact, intraoperative magnetic resonance imag
In conclusion, information-guided surgical management of ing guided system for conventional neurosurgical operating rooms.
gliomas using low-field-strength ioMRI based on the intrao- Neurosurgery 48:799 809
6. Ozawa N, Muragaki Y, Nakamura R, Iseki H (2008) Intraoperative
perative integration of anatomical, histopathological and diffusion weighted imaging for visualization of the pyramidal
neurophysiological data permits to perform maximal possible tracts. Part II: clinical study of usefulness and efficacy. Minim
tumor resection with minimal risk of postoperative neurolog- Invasive Neurosurg 51:67 71
ical morbidity, and may result in significant survival advan- 7. Senft C, Seifert V, Hermann E, Franz K, Gasser T (2008) Useful
ness of intraoperative ultra low field magnetic resonance imaging
tage, particularly in patients with WHO grade III gliomas. in glioma surgery. Neurosurgery 63(suppl 2):257 267
8. Ojemann JG, Miller JW, Silbergeld DL (1996) Preserved function
Conflict of interest statement We declare that we have no conflict of in brain invaded by tumor. Neurosurgery 39:253 259
interest. 9. Stummer W, Reulen HJ, Meinel T, Pichlmeier U, Schumacher W,
Tonn JC, Rohde V, Oppel F, Turowski B, Woiciechowsky C, Franz K,
Acknowledgements The authors are thankful to Drs. Osami Kubo, Pietsch T, ALA Glioma Study Group (2008) Extent of resection
Ken’ichi Hirasawa, Takemasa Kawamoto, Kosaku Amano, Yuichi and survival in glioblastoma multiforme: identification of and
Kubota, Tatsuya Ishikawa, Atsushi Watanabe, and Ayako Horiba adjustment for bias. Neurosurgery 62:564 576
(Department of Neurosurgery, Tokyo Women’s Medical University) 10. van den Bent MJ, Carpentier AF, Brandes AA, Sanson M,
for their help with the present study. This work was supported by Taphoorn MJ, Bernsen HJ, Frenay M, Tijssen CC, Grisold W,
the Industrial Technology Research Grant Program in 2000 2005 Sipos L, Haaxma Reiche H, Kros JM, van Kouwenhoven MC,
(A45003a) from the New Energy and Industrial Technology Develop Vecht CJ, Allqeier A, Lacombe D, Gorlia T (2006) Adjuvant
ment Organization of Japan (to Y. Muragaki). The research activities of procarbasine, lomustine, and vincristine improves progression
Drs. H. Iseki, M. Chernov, and K. Takakura are supported by the Program free survival but not overall survival in newly diagnosed anaplastic
for Promoting the Establishment of Strategic Research Centers, Special oligodendrogliomas and oligoastrocytomas: a randomized Europe
Coordination Funds for Promoting Science and Technology, Ministry of an Organisation for Research and Treatment of Cancer phase III
Education, Culture, Sports, Science and Technology (Japan). trial. J Clin Oncol 24:2715 2722
11. The Committee of Brain Tumor Registry of Japan (2003) Report of
brain tumor registry of Japan (1969 1996), 11th edition. Neurol
Med Chir (Tokyo) 43(suppl):1 111
References 12. Shinohara C, Muragaki Y, Maruyama T, Shimizu S, Tanaka M,
Kubota Y, Oikawa M, Nakamura R, Iseki H, Kubo O, Takakura K,
Hori T (2008) Long term prognostic assessment of 185 newly
1. Iseki H, Nakamura R, Muragaki Y, Suzuki T, Chernov M, Hori T,
diagnosed gliomas: Grade III glioma showed prognosis compara
Takakura K (2008) Advanced computer aided intraoperative tech
ble to that of Grade II glioma. Jpn J Clin Oncol 38:730 733
nologies for information guided surgical management of gliomas:
Implementation of the Ultra Low Field Intraoperative MRI
PoleStar N20 During Resection Control of Pituitary Adenomas
Ruediger Gerlach, Richard du Mesnil de Rochemont, Thomas Gasser, Gerhard Marquardt, Lioba Imoehl,
and Volker Seifert
Abstract Objective: To describe our experience with the macroadenomas. This system is of limited value for resection
application of an intraoperative ultra low field magnetic control of pituitary microadenomas.
resonance imaging system (ioMRI) PoleStar N20, Medtro-
nic Surgical Navigation Technologies, Louisville, USA dur- Keywords ioMRI Pituitary adenoma Suprasellar tumour
ing resection control of pituitary adenomas. Surgery
Methods: Forty-four patients were operated on a pituitary
adenoma (1 microadenoma, 43 macroadenomas; mean size
26.09.7 mm). The ioMRI system was used for navigation
and resection control after transseptal, transsphenoidal mi- Introduction
crosurgical tumour removal using standard instruments and
standard microscope. If any accessible tumour remnant was The estimation of tumour removal during microsurgical
suspected surgery was continued for navigation guided re- resection of pituitary adenomas can be intricate because
exploration and if necessary continued resection. of limited visualization of supra- and parasellar structures.
Results: The applications of the scanner integrated naviga- Especially the proof of adequate decompression of the optic
tion system, with a 3-planar reconstruction of the coronal pathway is crucial in patients with large suprasellar tumour
scan, enabled the surgeon to safely approach and remove the extension. To overcome this obvious problem intraoperative
tumour. The quality of preoperative tumour visualization MRI (ioMRI) resection control [1 9] has been propagated.
with the ultra low field ioMRI in patients with macroadeno- Intraoperative MRI systems differ with respect to scanner
mas is very good and has a good congruency with the preop- field strength (low field [1, 3, 4, 6, 8 15], high field [5, 7] or
erative 1.5 T MRI. For microadenomas the preoperative 3T [16, 17]) and therefore require diverse perquisites for
visualization is poor and very difficult to interpret. In seven implementation into a surgical procedure. According to the
patients ioMRI resection control showed residual tumours field strength of the various systems the operation room or at
leading to further resection. After final tumour resection the least the ioMRI system needs (ultra low field system Pole-
ioMRI scan documented adequate decompression of the Star N10 and N20) special shielding. To apply ioMRI either
optic pathway in all patients. However, the intraoperative a patient or scanner movement is necessary, which needs
image interpretation was equivocal in four patients in different technical installation requirements of the scanner
whom it was difficult to distinguish between small intrasellar and/ or operating table, which has implications for the oper-
tumour remnants and perioperative changes. ative work flow.
Conclusions: The PoleStar N20 is a safe, helpful and feasible It has been demonstrated that the rate of further resection
tool for navigation guided pituitary tumour approach. Image of non secreting pituitary adenomas was increased after the
interpretation is requires some experience, but decompression implementation of low [1, 4] and high field ioMRI scanner
of the optic system can be reliable shown in cases with pituitary systems [5, 7]. Furthermore the use of ioMRI increased the
rate of resection and therefore endocrinological cure of
patients with GH- secreting tumours [5].
R. Gerlach (*), R. du M. de Rochemont, T. Gasser, G. Marquardt, The PoleStar N20 (Medtronic Surgical Navigation Tech-
L. Imoehl, and V. Seifert nologies, Louisville, USA) is the second generation of a
Department of Neurosurgery and Neuroradiological Institute,
compact ultra low field ioMRI scanner. It represents a fur-
Johann Wolfgang Goethe University, Schleusenweg 2 16, 60528
Frankfurt/Main, Germany ther developed system, which was at first introduced by
e mail: r.gerlach@em.uni frankfurt.de Hadani et al. in 2001 [12].
74 R. Gerlach et al.
In this report we discuss our experience wit this ultra low

file intraoperative scanner type during microsurgical resec-
tion of pituitary adenomas.
Patients and Methods
The PoleStar N20 system was installed in the Department

of Neurosurgery, Johann Wolfgang Goethe- University
Frankfurt/ Main, Germany in September 2004. This scanner
type is applicable either with the use of a mobile shielding
device or in a shielded conventional operating room. In our
department the complete operating room was shielded.
We report about 44 patients (24 male, 20 female; mean
age 55.513.5 years) with pituitary macroadenomas (43)
and one patient with microadenoma, which were partly
included in a recent publication [11]. All patients had preop-
erative 1.5 T contrast enhanced MRI focused on the sellar
region to diagnose the extent of tumour growth and invasion
of the cavernous sinus.
A standard microsurgical procedure with a transseptal,
Fig. 1 Position of the patient within the magnet (PoleStar N20) from
transsphenoidal tumour resection was performed in all the patients left side. Because image quality strongly depends on a head
patients and the PoleStar N20 system was used for neurona- centered within the magnet positioning is a crucial step during the
vigation and preoperative tumour visualization and resection procedure. Shoulders are pulled down to optimize head position. For
control. All anaesthesiological equipment was fully ioMRI head fixation a 3 point sharp pin non ferromagnetic head holder was
used. The DRF is mounted on the head holder for navigation and the
compatible and used without any interference during the head coil is placed around the head
whole procedure. After introduction of general anaesthesia
and supine positioning the patients head was fixed in a
3-point pin titanium head clamp (Odin, Medical
Technologies Ltd; Yokneam, Israel) and a flexible surface
radiofrequency (rf) head coil is placed around the head. For
co-registration and navigation, a dynamic patient reference
frame (PRF) was fixed to the head holder (Fig. 1). During the
procedure the scanner is stored beneath the operating table.
For data acquisition the magnets were elevated into the scan-
position. To avoid any interfering during the scanning pro-
cedures and to achieve optimal image quality the power
supplies of all electrical medical devices, which are not
necessarily needed, were turned off during all scanning
procedures using a main switch. The system offers the ap-
plication of several T1 and T2 weighted sequences [14].
Selection of MRI sequences was described recently [11]
after these sequences were evaluated the ioMRI was used
with a defined protocol. The first scans were performed to
proof the correct head position in the centre of the magnet.
Therefore an axial ultra fast mixed T1/T2 weighted sequence
(esteady: TR 11.00 ms, TE 3.00 ms, 24 s, 8 mm, Fig. 2) or a Fig. 2 Short ioMRI sequence used for head positioning control using
short T1 sequence (T1: TR 40.0 ms, TE 3.00 ms, 60 s, 8 mm) an axial mixed T1/T2 sequence (esteady 24 s) showing the large intra
was used depending on the surgeons preference. To obtain and suprasellar tumour. Identification of correct head position is crucial
for the image quality and can be performed in axial or coronal plane
best image quality it is crucial to position the patients head in
the centre of the magnets. Moreover, for navigation purposes
the field of view (FOV) should cover the nasal cavity besides
Implementation of the Ultra Low Field Intraoperative MRI PoleStar N20 During Resection Control of Pituitary Adenomas 75
the sellar region to facilitate reliable navigation during the Technologies, Louisville, USA). A standard surgical micro-
transsphenoidal approach. scope (NC4, Karl Zeiss Medical, Oberkochen, Germany)
A coronal T1 weighted (TR 60.00 ms, TE 3.00 ms, 6.5 min, and normal ferro magnetic instruments were used through-
3 mm) gadolinium-DTPA enhanced (0.4 ml Gd/kg body, out the whole procedure.
Magnevist, Schering, Berlin; Germany) sequence was used When the surgeon considered the tumour to be completely
for navigation with calculation of a 3-planar reconstruction removed or at least achieved the intended amount of tumour
(Fig. 3). After the pre- resection scans the operative field was resection if the tumour had cavernous sinus invasion, a re-
prepared and sterile draped. To cover the magnet sterile section control scan was performed. The time from stopping
plastic drapes were used. For Navigation we used the Odin surgery to start ioMRI resection control scanning ranges
navigation software during the first time and later on the between 1 and 2 min. If an accessible tumour remnant was
stealth station (Vers.4.0; Medtronic Surgical Navigation visible on the ioMRI scan navigation guided surgical resec-
Fig. 3 A coronal T1 weighted Gd enhanced scan (3 mm) is used for neuronavigation. A triplanar reconstruction is calculated and displayed by the
system. (a) The Odin system was used during the first 8 months, (b) the Stealth station which was used afterwards and offers a virtual tip extension
for surgical guidance after application the navigation probe (c)
Fig. 4 Pre (left) and post resection (right) ultra low field ioMRI scan with 3 planar reconstruction for navigation guided (re ) exploration. The
complete removal of the suprasellar tumour with adequate decompression of the optic system is demonstrated, while the interpretation of the
intrasellar region is difficult due to oozing during obtaining the scan
tion was continued until further scan(s) documented com- noma. No adverse events related to the application of the
plete resection or the accomplished intended resection goal intraoperative ioMRI system PoleStar N20 were observed.
(Fig. 4). There were no accidents caused by the use of standard
ferromagnetic instruments. No procedure related complica-
tions occurred, especially no infection or surgery related
visual disturbance, or injury to the carotid artery.
Results Intraoperative imaging allowed an accurate localization
of the sellar region and pituitary tumours as well as identifi-
Forty-four patients underwent transsphenoidal tumour resec- cation of pertinent parasellar structures, which was the pre-
tion with the use of the ioMRI system PoleStar N20. Forty- requisite for safe and effective navigation guided resection
three patients had a macro- and one patient had a microade- (Fig. 3). Thus the navigation safely guided the surgeon
providing detailed anatomic information, which was also from GH- secreting adenomas, including the one patient
possible for the two patients with previous surgery and with continued surgery. Compared with the 3 months 1.5 T
operations for recurrent tumours. MRI a discrepancy in the images was found in these four
PoleStar N20 resection controlled surgery was feasible in cases were no clear residual intrasellar tumour was seen on
all patients with macroadenomas, while the visualization of ioMRI by the surgeon. Intraoperative interpretation was
small pituitary adenoma was limited and therefore patients challenging and a tiny small remnant could not be distin-
with microadenomas were not further evaluated. guished from perioperative changes but was identified in
According to the ioMRI surgery was continued in a total 1.5 T MRI at 3 months. However, re-evaluation for compar-
of 7 out of all 44 patients (15.9%). A continued resection ison analysis of the ioMRI scans depicted residual tumour
was possible in 1 of 15 patients (6.7%) with intended com- comparable to the 1.5 T MRI at the time of surgery. There-
plete resection, but the majority of cases were patients fore, this misinterpretation of the images was rather a prob-
with cavernous sinus invasion and therefore intended incom- lem of the surgeons, rather than the system itself.
plete resection. Thus the PoleStar N20 was mainly helpful
in detecting residual tumour in invasive tumour types,
were the surgeons tend to be less aggressive during tumour
resection. Discussion
The ophthalmologic status improved in all patients with
visual impairment (n¼19) and bilateral hemianopia (n¼18) Although there are associated financial burden the distribu-
except for one patients with blindness on one eye deriving tion and implementation of intraoperative MRI systems has
from previous surgery in another institution. Only one gained more and more acceptance. One of the most appre-
patient with intended incomplete resection and sticky ciated indications for ioMRI is the resection control of
tumour had a transient VIth nerve palsy, which completely pituitary macroadenomas (Fig. 5).
recovered within 6 weeks after surgery. In four of eight Two different developmental trends of intraoperative MRI
patients with GH-secreting tumours and one patient with systems have emerged over the last years. The implementation
prolactinoma endocrinological cure was achieved. Three of high field ioMRI systems on the one side offers the quality
patients with acromegaly had very small tumour remnants of diagnostic scanners during the operation, but has more
at 3 months 1.5 T MRI, which could not be shown by ioMRI. limitations in the surgical work flow. On the other end the
One patient had no visible tumour in standard control MRI 3 ultra low field system, PoleStar N20, enables the surgeon to
months postoperatively, but slightly elevated IgF 1 levels integrate the system with only minimal changes in the surgi-
thus consensus criteria were not met in this patient. cal work flow, but offers the possibility to use standard
Intraoperative image interpretation was equivocal in four instruments and microscope during the whole operation.
patients and unfortunately three of the four patients suffered We report our experience with an compact ultra low field
Fig. 5 Coronal T1 weighted

contrast enhanced [0.15 T] scan
before (left) and after (right)
resection of a large intra and
suprasellar adenoma (same
patient as in Fig. 2). The
intraoperative scan after resection
demonstrates resection of the
pituitary adenoma and a very
good visualization of the
suprasellar compartment with
adequate decompression of the
optic system and the pituitary
stalk with a deviation to the left
side is seen on the ultra low field
ioMRI PoleSTar system
scanner, which is positioned underneath the operating table Also, for small microadenomas the image quality has
during surgery and can be easily and fast moved up for clear limitations and the benefit of intraoperative MRI scan-
scanning purposes. This is the second generation of the ning in those patients is low.
compact ultra low field systems which demonstrates many IoMRI resection control of pituitary macroadenomas sig-
refinements compared to the first scanner of this type, which nificantly increases operation and total anaesthesia time, but
was introduce by Hadani et al. [12]. A larger gap between the the information of resection status and the chance for imme-
magnets eases patient positioning and the slightly higher field diate re-exploration and continued resection outweighs the
strength (0.15 T vs. 0.12 T) lead to improved image quality. longer and operation time [11, 13]. The reliable assessment
Similar to the N10 system no patient movement is neces- of the amount of resection increases the patients comfort and
sary with the PoleStar N20 system when a resection control avoids any uncertainty about the grade of resection, which
scan is required from the surgeon. Based on updated images usually exists postoperatively up to 3 months until a routine
acquired within minutes while surgery is paused remaining 1.5 T MRI documents the actual extend of resection.
tumour can be removed by further navigation- guided resec-
tion. The ioMRI navigation was accurate as described by
other groups [14, 18] and reliable to guide the surgeon during
the procedure. Thus, beside a safe tumour approach also in Conclusion
patients with residual tumour the use of the PoleStar N20
integrated navigation system allows a precise intraoperative
The 0.15 T ultra low field ioMRI PoleStar N20 offers imme-
tracking of residual tumour.
diate reliable resection control for suprasellar tumour parts,
However, the key question for ioMRI in pituitary surgery
while image interpretation of intra- and parasellar tumours is
is whether or not the implementation of an ioMRI increases
difficult. Therefore the correct image interpretation of intra-
the amount of tumour resection and therefore the number of
sellar compartment has an individually learning curve. After
complete tumour removal, which is the precondition for
implementation of the PoleStar N20 System continued
endocrinological cure in patients with secreting tumours.
resection increased the amount of resected tumour especially
Although not proven by randomized controlled trials the
in patients with intended subtotal tumour removal due to
use of an ioMRI improved the rate of complete resection
invasion of the cavernos sinus. The integrated navigation
independent of the ioMRI system. This was demonstrated
system displays accurate anatomic information also in
for both low and high field systems [1 8, 12, 14, 19].
patients with previous surgery.
Beside the above described advantages of the PoleStar
N20 system there are also drawbacks associated with its use. Disclosure None of the authors has any financial interest in the
This scanner type has an individually learning curve for methodology being advanced with the publication of our data. TG
image interpretation. The differentiation of small intrasellar serves as a clinical consultant for Medtronic Surgical Navigation Tech
nologies, Louisville, USA.
tumour remnants can be very difficult as described in a
recent series of patients with macroadenomas [11].
This uncertainty of intraoperative interpretation by the
surgeon was due to difficulties in interpretation rather than
a methodological problem of the ioMRI system and confirms References
that a high level of experience is necessary for adequate
image interpretation. A comparative post surgery analysis 1. Bohinski RJ, Warnick RE, Gaskill Shipley MF, Zuccarello M, van
comparing the intraoperative ultralow filed MR-images and Loveren HR, Kormos DW, Tew JM Jr (2001) Intraoperative mag
netic resonance imaging to determine the extent of resection of
the 1.5 T standard 3 months images showed these small pituitary macroadenomas during transsphenoidal microsurgery.
residual tumours in the ultra low field obtained images. Neurosurgery 49:1133 1143
The image quality was significantly better for the suprasellar 2. Darakchiev BJ, Tew JM Jr, Bohinski RJ, Warnick RE (2005)
compartment compared to the intra- and parasellar compart- Adaptation of a standard low field (0.3 T) system to the operating
room: focus on pituitary adenomas. Neurosurg Clin N Am
ment, which is important to assess the adequate decompression 16:155 164
of the optic pathway and complete removal of suprasellar 3. De Witte O, Makiese O, Wikler D, Levivier M, Vandensteene A,
tumour parts. This is in accordance with the experience of Pandin P, Baleriaux D, Brotchi J (2005) Transsphenoidal approach
other groups using various low field systems [1, 4, 6, 14, 19, with low field MRI for pituitary adenoma. Neurochirurgie
51:577 583
20]. Sufficient decompression of the chiasm is of paramount 4. Fahlbusch R, Ganslandt O, Buchfelder M, Schott W, Nimsky C
significance especially in patients with large invasive para- (2001) Intraoperative magnetic resonance imaging during trans
sellar tumours, not accessible for complete resection. sphenoidal surgery. J Neurosurg 95:381 390
5. Fahlbusch R, Keller B, Ganslandt O, Kreutzer J, Nimsky C (2005) ing guided system for conventional neurosurgical operating rooms.
Transsphenoidal surgery in acromegaly investigated by intraopera Neurosurgery 48:799 807
tive high field magnetic resonance imaging. Eur J Endocrinol 13. Kanner AA, Vogelbaum MA, Mayberg MR, Weisenberger JP,
153:239 248 Barnett GH (2002) Intracranial navigation by using low field
6. Martin CH, Schwartz R, Jolesz F, Black PM (1999) Transsphenoi intraoperative magnetic resonance imaging: preliminary experi
dal resection of pituitary adenomas in an intraoperative MRI unit. ence. J Neurosurg 97:1115 1124
Pituitary 2:155 162 14. Schulder M, Salas S, Brimacombe M, Fine P, Catrambone J,
7. Nimsky C, von Keller B, Ganslandt O, Fahlbusch R (2006) Intrao Maniker AH, Carmel PW (2006) Cranial surgery with an expanded
perative high field magnetic resonance imaging in transsphenoidal compact intraoperative magnetic resonance imager. Technical
surgery of hormonally inactive pituitary macroadenomas. Neuro note. J Neurosurg 104:611 617
surgery 59:105 114 15. Wu JS, Shou XF, Yao CJ, Wang YF, Zhuang DX, Mao Y,
8. Schwartz TH, Stieg PE, Anand VK (2006) Endoscopic transsphe Li SQ, Zhou LF (2009) Transsphenoidal pituitary macroadenomas
noidal pituitary surgery with intraoperative magnetic resonance resection guided by PoleStar N20 low field intraoperative magnet
imaging. Neurosurgery 58:ONS44 ONS51 ic resonance imaging: comparison with early postoperative high
9. Steinmeier R, Fahlbusch R, Ganslandt O, Nimsky C, Buchfelder M, field magnetic resonance imaging. Neurosurgery 65:63 70
Kaus M, Heigl T, Lenz G, Kuth R, Huk W (1998) Intraoperative 16. Hall WA, Galicich W, Bergman T, Truwit CL (2006) 3 Tesla
magnetic resonance imaging with the magnetom open scanner: intraoperative MR imaging for neurosurgery. J Neurooncol
concepts, neurosurgical indications, and procedures: a preliminary 77:297 303
report. Neurosurgery 43:739 747 17. Pamir MN, Peker S, Ozek MM, Dincer A (2006) Intraoperative
10. Black PM, Moriarty T, Alexander E III, Stieg P, Woodard EJ, MR imaging: preliminary results with 3 tesla MR system. Acta
Gleason PL, Martin CH, Kikinis R, Schwartz RB, Jolesz FA Neurochir Suppl 98:97 100
(1997) Development and implementation of intraoperative mag 18. Salas S, Brimacombe M, Schulder M (2007) Stereotactic accuracy
netic resonance imaging and its neurosurgical applications. Neuro of a compact intraoperative MRI system. Stereotact Funct Neuro
surgery 41:831 842 surg 85:69 74
11. Gerlach R, de RR Du Mesnil, Gasser T, Marquardt G, Reusch J, 19. Pergolizzi RS Jr, Nabavi A, Schwartz RB, Hsu L, Wong TZ,
Imoehl L, Seifert V (2008) Feasibility of Polestar N20, an ultra Martin C, Black PM, Jolesz FA (2001) Intra operative MR guid
low field intraoperative magnetic resonance imaging system in ance during trans sphenoidal pituitary resection: preliminary
resection control of pituitary macroadenomas: lessons learned results. J Magn Reson Imaging 13:136 141
from the first 40 cases. Neurosurgery 63:272 284 20. Schulder M, Sernas TJ, Carmel PW (2003) Cranial surgery and
12. Hadani M, Spiegelman R, Feldman Z, Berkenstadt H, Ram Z navigation with a compact intraoperative MRI system. Acta
(2001) Novel, compact, intraoperative magnetic resonance imag Neurochir Suppl 85:79 86
Intraoperative MRI for Stereotactic Biopsy
Michael Schulder and David Spiro
Abstract This work aims at demonstrating the value of Introduction

intraoperative magnetic resonance imaging (ioMRI) as a
routine tool for stereotactic brain biopsy. Techniques for brain biopsy have evolved in relation to the
Biopsies were done using the PoleStar N-20 ioMRI evolution of imaging methods. Until the advent of computed
(Medtronic Navigation, Louisville, Colorado, USA) under tomography (CT) stereotactic techniques, developed primar-
general anesthesia. Images were acquired after patient posi- ily for functional neurosurgical operations, were not appli-
tioning and after insertion of an MRI-compatible biopsy cable for morphologic targets except in rare cases. The
cannula. A Navigus guide (Medtronic Navigation) was advent of CT allowed for free-hand biopsies to be done
used to align and direct the cannula. Retargeting was done with imaging control, but this approach lacked the ability
as necessary, to improve placement within the target and to for any stereotactic localization (except in the plane of a
avoid critical structures, using the system’s integrated infra- particular image) [1, 2]. The joining of CT to stereotactic
red navigation tool. Cannula placement was tracked using frames in the mid-1980s allowed for the emergence of ste-
serial images. reotactic biopsy as a routine procedure [3, 4]. Several years
ioMRI-guided biopsy was done in 39 patients, of whom later magnetic resonance imaging (MRI) became available
28 had neoplasms and 11 had non-neoplastic conditions. for this use. More recently, surgical navigation technology
Additional OR time related to the use of ioMRI (including (‘‘frameless stereotaxy’’) has become widely accepted. Com-
positioning of the patient and magnet, and imaging acquisi- parison of stereotactic biopsy with a frame or frameless
tion) averaged 1.1 h. In 53% of the surgeries the biopsy approach has shown equivalent efficacy and safety between
cannula was repositioned based on intraoperative imaging. these techniques.
A histologic diagnosis was obtained in all but one patient, Intraoperative MRI (ioMRI) was introduced by Black
with ioMRI confirming proper cannula placement in all et al. at the Brigham and Women’s Hospital in the mid-
cases. There were no significant hemorrhages on clinical or 1990s [5]. The impetus behind this technological leap was
imaging grounds nor any other complications. to provide updated images that would account for the
IoMRI can be routinely used for stereotactic biopsy in a inevitable brain shift that renders information from preop-
regular neurosurgical operating environment. While general erative images unreliable for safe navigation. In addition,
anesthesia is used and there is some additional time incurred images can of course be used for resection control and to
from this technology the improved diagnostic yield and rule out such complications as hemorrhage or stroke. The
ability to avoid complications make ioMRI an ideal techni- logic of adapting ioMRI for stereotactic biopsy lies in the
cal adjunct for brain biopsy. ability to ensure accurate and safe placement of the sam-
pling instrument, and to in essence make it an image
Keywords Frameless stereotaxy Intraoperative magnetic controlled and not ‘‘blind’’ surgery. However, the effort
resonance imaging Stereotactic brain biopsy and expense of using ioMRI for stereotactic biopsy a
procedure that is viewed (with reasonable accuracy) as
routine and safe may fairly be questioned.
We describe our experience with the PoleStar N20
(Medtronic Navigation, Louisville, Colorado, USA), a
M. Schulder (*) and D. Spiro low field strength ioMRI, designed for use in a regular
Department of Neurosurgery, North Shore LIJ, North Shore University
Hospital, 9 Tower, Manhasset, NY 11030, USA neurosurgical operating room (OR), for stereotactic brain
e mail: Schulder@nshs.edu biopsy.
82 M. Schulder and D. Spiro
Surgical Technique After acquiring a 6.5 min T1 weighted image (with con-
trast as needed), a surgical plan is made using the desired
Patients selected for ioMRI-guided biopsy routinely undergo target and entry points. If the surgeon wishes, the image may
stereotactic MRI in advance of surgery. This allows confir- be exported to the StealthStation software, which resides as
mation that there has been no change in the lesion, for a separate program on the PoleStar computer. An incision
patients in whom the time between imaging and surgery long enough to accommodate the Navigus biopsy guide
is prolonged, and also provides a ‘‘hedge’’ should there be (Medtronic Navigation) is made, about 4 cm long. The burr
any technical difficulties with the ioMRI unit. In such an hole and dural opening are done before seating the Navigus.
event surgery can proceed using ‘‘conventional’’ surgical The dedicated trackable probe is placed in the Navigus and
navigation. aligned with the planned trajectory. A side-cutting MRI-
The PoleStar N20 ioMRI has been described in previous compatible biopsy cannula is stopped at the appropriate
reports. It is the second generation of an innovative device length and inserted towards the target. The magnet is raised
designed by Hadani and colleagues [6, 7]. In brief, it is based and short image sequences, between 24 s and 1 min, are
on a mobile, 0.15 Tesla (T) permanent magnet. The system acquired. If cannula position is adequate, then biopsies are
is stored in a cage in the back of the OR and brought out taken at 2 3 levels if possible. A frozen section is requested.
when needed. The OR is shielded against electrical interfer- The biopsy cannula is redirected if intraoperative imaging
ence from outside sources. During imaging, non-filtered shows that it is not adequately placed within the lesion, or is
electrical devices are shut off and unplugged. Anesthesia too close to a critical structure such as a blood vessel.
equipment is MRI-compatible. The magnet poles are sepa- Imaging is repeated after each redirection. After taking the
rated by a 27 cm gap. An integrated infrared navigation tool last specimen and removing the cannula a last image is
is included in the system. The magnet is positioned under the acquired to rule out a hematoma.
OR table, raised as needed for imaging, and lowered again at
other times. The low magnet field strength allows for the use
of regular operating tools, including ferromagnetic ones, as
long as the magnet poles are below the OR table. Patient Data
General anesthesia is used, as patients require three-point
head fixation in the MRI-compatible headholder that is part IoMRI-guided stereotactic biopsies have been done on 39
of the PoleStar system. Patients are positioned on a regular patients. Their age ranged from 5 years to 86 years, with a
OR table. In our experience the prone position should be mean of 49 years. There were 25 male and 14 female
used when biopsy targets are located in the posterior half of patients. Lesion locations, as in other series, reflected
the intracranial space; lateral positioning is possible but a typical distribution in the brain, including frontal [8],
somewhat more cumbersome. A disposable, flexible MR parietal [8], temporal [9], and occipital [10] lobes. Two
receiving coil is positioned and draped out of the field; this patients had pineal lesions, two in the thalamus, and one
avoids the need for coil replacement during surgery (Fig. 1). each in the corpus callosum, suprasellar region, midbrain,
Fig. 1 Patient positioned prone

for right parietal stereotactic
biopsy in the PoleStar N20. Note
that the flexible receive coil is
positioned to allow for it to be
draped out of the field. The
reference frame for the infrared
navigation tool is to the patient’s
left
Intraoperative MRI for Stereotactic Biopsy 83
hypothalamus, and cerebellum. The dominant hemisphere cocci. The patient was treated with antibiotics and the lesion
contained the target in 18 patients. The supine position was regressed.
used in 28 patients, while 11 were turned prone. Patient 2. This 70 year old right handed man presented
with 2 weeks of left hemiparesis. Imaging revealed a ring-
enhancing mass in the right insular cortex, with some
involvement of the right Sylvian fissure. The medial aspect
Results of the lesion was targeted to avoid injury to Sylvian vessels
(Fig. 3a). Intraoperative imaging resulted in improved can-
On one occasion the PoleStar N20 did not properly boot. As nula placement (Fig. 3b, c). The patient proved to have a
the patient had been referred specifically for ioMRI-guided high grade glioma.
biopsy, having had a nondiagnostic biopsy elsewhere, the
procedure was aborted and rescheduled. The number of
scanning sessions ranged from 2 to 9 (including multiple
images as biopsies were obtained), with a mean of 3.5 per
procedure. In 21 cases (53%) the intraoperative image led to
Discussion
replacement of the biopsy cannula. In 12 patients this change
was made to improve cannula position in the target; in 4 The progress of neurosurgery over the last century has been
patients, to avoid injury to a critical structure before taking a marked by decreasing the need for guesswork. The practical
biopsy; and in 5 after an initial frozen section did not obvi- beginning of minimally invasive approaches for brain biopsy
ously show lesional tissue. The additional time needed for came in the 1980s with the performance of freehand proce-
setup and use of ioMRI ranged from 0.5 to 2.0 h, with a men dures done in CT scanners. A review of several CT-guided
of 1.1 h. No significant hemorrhage was detected on final freehand biopsy series showed a diagnostic yield of 90%
intraoperative images. All patients were neurologically sta- (range 79 97%) a mortality rate of 2.5% (0.5 4.7%) and
ble after surgery. There were two deaths in the series due to morbidity rate of 7.8% [2]. For obvious technical reasons,
progression of pre-existing disease lymphoma in one and these procedures were best reserved for patients with rela-
progressive multifocal leukoencephalopathy from AIDS in tively large and superficial lesions.
the other. The stereotactic frame, invented by Horsley and Clarke
Diagnoses included high grade glioma in 15 patients, low in 1908, was first adapted for human surgery in 1947 by
grade glioma in 11, lymphoma in 1, and embryonal cell Spiegel and Wycis [9]. For nearly 40 years thereafter, frames
carcinoma in 1; and 10 patients had non-neoplastic lesions, were used for functional neurosurgery, using ventriculo-
including 3 bacterial abscesses. In one patient, although graphy for guidance. The combination of the stereotactic
increased cellularity was seen, no definitive diagnosis was concept with CT and then MRI allowed for surgery aimed
made. Intraoperative images showed the biopsy cannula at morphologic targets. Interestingly, the published experi-
to be within the enhancing component of the lesion (this ence with stereotactic biopsies yielded results that were
patient remains neurologically intact, with stable MRI find- grossly similar to those achieved with freehand approaches
ings, 1 year later). (although these series included targets that were smaller and
deeper than those attempted before) [2]. With the advent of
surgical navigation, sometimes called ‘‘frameless stereo-
taxy’’, similar results have been obtained a diagnostic
Case Illustrations rate of slightly over 90%, with morbidity and mortality of
0.7 % and 3.5% [11].
Patient 1. This 44 year old right handed man was transferred Why then bother introducing a new technique for stereo-
from another hospital, having come to the emergency room tactic biopsy? The answer lies in the continued incidence of
complaining of progressive headache and then left sided non-diagnostic procedures and the complications, which can
weakness. MRI showed a ring-enhancing lesion in the right reflect unrecognized intracranial hemorrhage, injury to blood
thalamus, with some surrounding edema (Fig. 2a); symp- vessels, or unnecessary entry into critical brain areas. Intrao-
toms improved on steroids. The preoperative diagnosis was perative imaging offers several advantages in this regard (1)
high grade glioma. The patient was positioned supine for brain shifts that may occur with dural opening can be
biopsy and a right coronal entry point was chosen (Fig. 2b). accounted for; (2) if a non-lesional frozen section is obtained,
Imaging showed the biopsy cannula to be at the enhancing repeat imaging and targeting can be done during the same
rim (Fig. 2c). Frozen section showed inflammatory tissue, surgery; (3) the surgeon can ensure that the biopsy is taken
and specimens for permanent pathology. Cultures were neg- from the area most likely to yield a diagnosis; (4) with serial
ative but pathology confirmed an abscess with gram positive imaging during biopsy, unnecessary penetrations with the
Fig. 2 (a) Preoperative ioMRI of

a 44 year old man thought to have
a right thalamic glioma. (b)
Reconstructed triplanar views
showing selected target on
enhancing rim. (c) 24 s T2 ioMRI
shows cannula at the target point.
The lesion proved to be a
bacterial abscess
Fig. 3 70 year old man with

mass in the right insula. (a)
Targeting of the medial rim of the
tumor, away from Sylvian
vessels. (b) Image acquired after
first pass of cannula. (c) Improved
position of cannula to increase
diagnostic yield, still avoiding
Sylvian fissure. Frozen section
and permanent pathology showed
glioblastoma
cannula can be avoided; (5) hemorrhage can be ruled out at Other objections that may be raised to this ioMRI-guided
the end of the procedure. biopsy technique are the need for a bur hole and the longer
For the foreseeable future brain imaging, intraoperatively incision needed to accommodate the Navigus guide. This
or otherwise, will be best done with MRI. IoMRI-guided is in contrast to frameless and especially frame-based
biopsy was first described by Black et al., using a 0.5 T approaches where a 5 mm incision and twist drill hole may
magnet, in their pioneering article [5]. The authors reported suffice. In addition, the Navigus is a disposable device,
a 100% diagnostic yield for brain biopsy; one patient was whose use brings an added expense to the surgery. Finally,
found to have an intracranial hemorrhage that required although intravenous sedation and local anesthetic may be
surgical evacuation. Bernays et al. reported a series devoted considered for these cases, the need for 3-point head fixation
specifically to stereotactic biopsy, using a specially- and the added time of ioMRI use make general anesthesia far
designed, skull-mounted guide. In 113/114 patients a diag- easier. This is a potential added source of morbidity com-
nosis was made; morbidity was 1.8% and mortality 0.9% pared to frame-based biopsies, which typically are done
[10]. Hall et al. performed biopsies in a diagnostic 1.5 MRI under local anesthesia.
suite, modified for surgical use. The first 35 procedures were
done freehand, and the following 40 with a device that
evolved into the Navigus guide. The histological yield was
100%, with morbidity of 2.8% and mortality of 1.4% [12]. Conclusions
Ours is the first series to focus on use of a compact, low-
field ioMRI for stereotactic biopsy. Several small series from
Stereotactic brain biopsy can be done as a routine procedure
groups using 0.2 T ioMRIs included diagnostic rates of
in a low field ioMRI environment. Imaging enables the
93 100%, with limited description of technique [8, 13, 14].
surgeon to compensate for brain shift, to re-image and re-
Kanner et al. reported their use of the PoleStar N10, which
target as necessary, and to rule out intracranial bleeding.
used a 0.12 T magnet [15]. Their overall experience included
With the increasing adoption of ioMRI at neurosurgical
biopsies in 15 patients, all of which were diagnostic. As noted
centers worldwide, we recommend that this be the method
above, we made a histological diagnosis in all but one of our
of choice for stereotactic biopsy.
patients, but were able to confirm that the biopsy was taken
from the area of interest. In addition, in this and in four other
Conflicts of interest statement We declare that we have no conflict
patients, the ability to image the cannula after receiving a of interest.
non-lesional frozen section report allowed for retargeting as
needed, without ‘‘poking the brain’’ based merely on a pre-
operative image. While additional time (mean 1.1 h) was
incurred with the use of ioMRI, this does not take into
References
account the time saved by avoiding placement of a stereotac-
tic frame and the subsequent imaging session in a diagnostic
1. Greenblatt SH, Rayport M, Savolaine ER, Harris JH, Hitchins MW
scanner. Additional time could be saved by eliminating the (1982) Computed tomography guided intracranial biopsy and cyst
frozen section and relying on the images alone. We feel that aspiration. Neurosurgery 11:589 598
the advantages of intraoperative pathology, which has been 2. Wen DY, Hall WA, Miller DA, Seljeskog EL, Maxwell RE (1993)
shown to increase the diagnostic yield of stereotactic biop- Targeted brain biopsy: a comparison of freehand computed
tomography guided and stereotactic techniques. Neurosurgery
sies, outweigh this small prolongation of surgery [16]. 32:407 412; discussion 412 413
Advocates for the high field ioMRI approach note the 3. Gildenberg PL (1983) Stereotactic neurosurgery and computerized
improved resolution and the ability to acquire diffusion tomographic scanning. Appl Neurophysiol 46:170 179
weighted imaging, MR spectroscopy or angiography, or 4. Heilbrun MP (1983) Computed tomography guided stereotactic
systems. Clin Neurosurg 31:564 581
functional images with these systems. It is not clear that 5. Black P, Moriarty T, Alexander E 3rd, Stieg P, Woodard E,
these advanced capabilities, whatever their worth may be, Gleason P, Martin C, Kikinis R, Schwartz R, Jolesz F (1997)
would be required to perform what should be a routine Development and implementation of intraoperative magnetic reso
procedure, namely, stereotactic biopsy. Compared to these nance imaging and its neurosurgical applications. Neurosurgery
41:831 842; discussion 842 835
other ioMRI devices, the PoleStar N20 offers some distinct 6. Hadani M, Spiegelman R, Feldman Z, Berkenstadt H, Ram Z
advantages. These include the use of a regular OR table in a (2001) Novel, compact, intraoperative magnetic resonance imaging
regular OR. This allows for patient positioning as dictated by guided system for conventional neurosurgical operating rooms.
the surgical approach, and causes minimal disruption to Neurosurgery 48:799 807
7. Schulder M, Salas S, Brimacombe M, Fine P, Catrambone J,
surgical routine. However, it must be noted that the incom- Maniker AH, Carmel PW (2006) Cranial surgery with an expanded
plete field of view of the PoleStar N20 may limit imaging of compact intraoperative magnetic resonance imager. Technical
targets in the posterior fossa. note. J Neurosurg 104:611 617
8. Tronnier VM, Wirtz CR, Knauth M, Lenz G, Pastyr O, 13. Bernstein M, Al Anazi AR, Kucharczyk W, Manninen P, Bronskill M,
Bonsanto MM, Albert FK, Kuth R, Staubert A, Schlegel W, Henkelman M (2000) Brain tumor surgery with the Toronto open
Sartor K, Kunze S (1997) Intraoperative diagnostic and interven magnetic resonance imaging system: preliminary results for 36
tional magnetic resonance imaging in neurosurgery. Neurosurgery patients and analysis of advantages, disadvantages, and future
40:891 900; discussion 900 902 prospects. Neurosurgery 46:900 907; discussion 907 909
9. Gildenberg PL (1990) The history of stereotactic neurosurgery. 14. Fahlbusch R, Ganslandt O, Nimsky C (2000) Intraoperative imag
Neurosurg Clin N Am 1:765 780 ing with open magnetic resonance imaging and neuronavigation.
10. Bernays RL, Kollias SS, Khan N, Brandner S, Meier S, Yonekawa Y Childs Nerv Syst 16:829 831
(2002) Histological yield, complications, and technological con 15. Kanner AA, Vogelbaum MA, Mayberg MR, Weisenberger JP,
siderations in 114 consecutive frameless stereotactic biopsy pro Barnett GH (2002) Intracranial navigation by using low field
cedures aided by open intraoperative magnetic resonance imaging. intraoperative magnetic resonance imaging: preliminary experi
J Neurosurg 97:354 362 ence. J Neurosurg 97:1115 1124
11. Woodworth GF, McGirt MJ, Samdani A, Garonzik I, Olivi A, 16. Kim JE, Kim DG, Paek SH, Jung HW (2003) Stereotactic biopsy
Weingart JD (2006) Frameless image guided stereotactic brain for intracranial lesions: reliability and its impact on the planning
biopsy procedure: diagnostic yield, surgical morbidity, and com of treatment. Acta Neurochir (Wien) 145:547 554; discussion
parison with the frame based technique. J Neurosurg 104:233 237 554 555
12. Hall WA, Martin AJ, Liu H, Nussbaum ES, Maxwell RE,
Truwit CL (1999) Brain biopsy using high field strength interven
tional magnetic resonance imaging. Neurosurgery 44:807 813
The Evolution of ioMRI Utilization for Pediatric Neurosurgery: A
Single Center Experience
Thomas M. Moriarty and W. Lee Titsworth
Abstract From its inception intraoperative magnetic reso- Introduction

nance imaging (ioMRI) was envisioned to have significant
applications in neurosurgery in general and pediatrics spe- The first intraoperative MRI system (ioMRI) was a collabo-
cifically. Over the last 9 years we have noted a dramatic shift rative product between researchers at the Brigham and
in our ioMRI usage from intracranial tumors to cerebrospi- Women’s Hospital in Boston, MA, and General Electric
nal fluid management and complex cysts. Here we present Medical Systems first implemented in 1996 [1]. The first
seven selected cases to illustrate lessons learned from our ioMRI pediatric neurosurgery was performed on a small
operative experience within the GE Signa SP/I open-config- cerebellar tumor soon after. From its inception, ioMRI was
uration ‘‘double-doughnut’’ MRI. These cases including a seen as a tool most suited to neurosurgical application and
ganglioglioma, ependymoma, and pilocytic astrocytoma was anticipated to have particular utility in tumor, epilepsy,
tumor resection, as well as arachnoid cysts, complex cyst, hydrocephalus, cystic lesions, and potential for minimally
and microabscess drainage reflect our current use of ioMRI invasive surgery [1, 2]. This subset of cases predicted that
in pediatric neurosurgical cases. Namely that ioMRI is opti- ioMRI would hold particular value in pediatric neurosur-
mal for (1) resection of small tumors with poorly differen- gery. Here we present a review of the ioMRI usage from
tiated tumor margins, (2) large tumors with mass effect, and 2000 to 2009 from a pediatric neurosurgery perspective
(3) shunt or catheter placement requiring either extreme accu- using the GE double doughnut system.
racy or intraoperative confirmation of catheter placement. We There are currently three basic ioMRI concepts which can
also comment on the legitimate limitations of this technology be categorized as either low, mid, or high-field systems.
in certain operations. Additionally emphasized are cases in Low-field ioMRI consists of a 0.12-T magnet mounted direct-
which ioMRI imaging drives operative decision making, ly on the OR table. When not in use the two 40-cm diameter
highlighting the unique and unequaled abilities of this tech- discs, located 25 cm apart, are kept under the table. This
nology for a subset of pediatric neurosurgical cases. device, which was originally designed by Odin, is available
now through Medtronic as the Polestar [3]. While offering a
Keywords Catheter CSF diversion Ependymoma lower startup cost, this product has significantly poorer
Ganglioglioma Interventional MRI Intracranial abscesses image quality and limited MR capabilities. In contrast,
Intracranial tumors Neurosurgery Pediatric neurosurgery mid-field ioMRI, originally developed by GE, consists of a
Subarachnoid cyst ‘‘double-doughnut’’ design in which the surgeon operates
while imaging. The physical design utilizes two supercon-
ducting magnets spaced 58 cm apart. This system offers the
advantage of good quality images in a timely manner; how-
ever ergonomics and the requirement of special instrumen-
tation limit its application in some surgeries. High-field
systems utilize either a 1.5-T magnet that comes to and
T.M. Moriarty (*) from the patient or requires moving the patient in and out
Kosair Children’s Hospital, Norton Neuroscience Institute, 210 E Gray of a stationary magnet [2, 4, 5]. These systems offer the
St.Suite 1105, Louisville, KY 40202, USA
advantage of improved images and the use of ferromagnetic
e mail: Thomas.Moriarty@nortonhealthcare.org
instrumentation during portions of the operation but trans-
W.L. Titsworth
Department of Anatomical Sciences and Neurobiology, University of port time limits the speed of image acquisition. Some recent
Louisville School of Medicine, Louisville, KY 40202, USA versions of this configuration are the IMRIS, BrainSUITE,
90 T.M. Moriarty and W.L. Titsworth
GE Amiga, and the Phillips system. A more thorough dis- ioMRI Utilization in Tumors
cussion of each system can be found elsewhere [6]. While
pediatric neurosurgery has been performed on each of these During the initial days of our ioMRI usage we attempted a
configurations, this discussion will be limited to our experi- broad range of tumor operations within the ioMRI suit. Over
ence with the GE double doughnut ioMRI. time it became clear that ioMRI facilitated the removal of
supratentorial tumors but had ergonomic challenges in pos-
terior fossa tumors operations. However, it remained a good
tool for posterior fossa stereotaxy as the following cases will
Historical Use of ioMRI in One Institution illustrate. Traditionally low-grade gliomas are particularly
difficult to remove since neoplastic tissue so closely resem-
From 2000 to 2009 we utilized a Signa SP/I open- bles normal tissue in these lesions [7, 8]. However, patients
configuration ‘‘double-doughnut’’ MRI (GE Medical Systems, who underwent subtotal resection are 1.4 times more like to
Milwaukee, WI, USA). Our ioMRI suit is a fully functional have disease recurrence and have almost 5 times the risk of
operating room located in the adult surgical facility with death relative to patients who underwent gross total resec-
direct access to the pediatric hospital. Over the last 9 years tion [9]. In cases such as these, ioMRI is an ideal tool for
we have preformed 610 neurosurgical cases, 282 of which removal of all necessary malignancy while maximally
were pediatric neurosurgery, with a majority of cases being sparing native tissue. Case #1 is a 14 year old male who
craniotomies (44%), cyst management (25%), or CSF diver- presented with a seizure disorder and was subsequently
sion (16%). Within our pediatric caseload we noticed a shift found to have a right temporal low grade ganglioglioma
in usage over this time. During the first 2 years of service (WHO I) (Fig. 1a c). While preoperative T1 sequence
76% of pediatric cases were craniotomies for the removal of showed little abnormality, T2 images demonstrated a clear
masses while only 24% were either cyst drainage or CSF unifocal lesion suggesting the possibility of gross total
diversions. In contrast, over the last 2 years only 35% of resection [8]. Intraoperative T2 weighted sequences allowed
pediatric cases were craniotomies while 65% were now proper localization of the lesion and a clean resection of a
complex cysts or shunts. This shift occurred secondary to woody mass was achieved without disturbing normal tissue.
our recognition of this systems strengths and limitations. Repeat imaging at 2 years showed no residual tumor and the
Particularly we realized that not all tumors were ideally patient remained seizure free.
suited for ioMRI and secondly this technology is unrivaled In contrast to diminutive lesions, ioMRI also shows advan-
in targeting small foci because it provided stereotaxis with tages over static stereotaxy in large intracranial masses. Case
instant real-time feedback and no discordance between real- #2 is an 11 year old male with a 65.74.6 cm3 anaplastic
ity and the virtual construct. ependymoma (WHO III) with marked surrounding edema
a Case #1 b c
d e f
Fig. 1 Case #1: Low grade Case #2
Ganglioglioma. T2 weighted
preoperative (a) and
intraoperative (b, c) images. Note
the pointer in (b) used for
localization. Case #2: Anaplastic
ependymoma. Preoperative (d),
intraoperative (e), and 5 month
postoperative images (f). Note
significant surrounding edema
and mass effect (d)
The Evolution of ioMRI Utilization for Pediatric Neurosurgery: A Single Center Experience 91
(Fig. 1d f). Again subtotal resection of these tumors has been viewpoint. Due to the potential spontaneous regression of
shown to affects outcomes [10, 11]. ioMRI allowed access to these benign tumors the decision was made not to retrieve
the tumor via the most direct pathway similar to what is the unseen residual tumor and follow the patient with serial
achieved with static stereotaxy. Next the small cystic compo- scans, a treatment option which has been well described by
nent of the tumor was decompressed. However, frequent Hayward and colleges [12]. While we found posterior fossa
intraoperative imaging was used to quickly debulk the approaches particularly challenging in the ‘‘double dough-
tumor. We were then able to remove the lesion along a clear nut’’ system it should be noted that successful posterior fossa
plane. Five month postoperative images showed a barely operations have been achieved with other ioMRI designs
perceptible surgical cavity. Four and a half years later the that allow more surgical freedom, such as the low-field
patient remains free from recurrence. The challenge of this Polestar [13] and the high-field systems.
operation in regards to image guided surgery was twofold. However posterior fossa stereotactic procedures are
First, after craniotomy the shifting of intracranial contents unencumbered by the limited operative space within the
rendered all preoperative images useless. Secondly, the double doughnut since catheter or probe placement does
surrounding edema so distorted normal tissue so as to not require direct visualization by the surgeon. The advan-
obscure the tumor margins. Using ioMRI to quickly debulk tage of utilizing ioMRI in these cases is the continually
the tumor allowed decompression of the surrounding tissue updated image information allows more precision when
and brought into view a natural tissue plane. operating near eloquent anatomy of the brain stem. Case
However, this technology is not well suited to all tumors. #4 is a 5 year old female who presented with increasing
First, we have learned through experience that ioMRI holds problems piano playing, gait changes, vision changes, and
little advantage over the standard operating facilities for the mild facial droop. Subsequent imaging revealed a cystic
resection of tumors with clear margins. Additionally the astrocytoma of the midbrain. Chemotherapy was initiated
limited operative environment in the ‘‘double doughnut’’ however the enlarging cyst was causing obstructive hydro-
design actually hinders this systems application in posterior cephalus and worsening of symptoms. Therefore it was
fossa approaches. This is demonstrated by Case #3 in which decided to place a catheter through the right cerebellar
a 5 year old female who presented with a posterior fossa hemisphere into the cystic component using ioMRI. Follow-
pilocytic astrocytoma (Fig. 2a c). Due to the 58 cm of ing placement of the catheter and confirmation in three planes
operative space in the ‘‘double doughnut’’ system this ap- by ioMRI, 15 cc of fluid was removed and a Rickham reser-
proach can only be achieved by moving the prone patient voir was attached and secured. The patient then continued her
superiorly while the surgeon is forced to abut the inferior chemotherapy regimen without incident. Follow-up imaging
magnet (see Fig. 2d, e). In this case, even with repositioning, showed incremental improvements until no sign of tumor or
the most superior portion of the tumor, which was clearly enhancement remained at 2 years postoperatively. She has
visible on ioMRI, remained unseen from the surgeon’s remained free from recurrence for 8 years.
a b c
Case #3
Fig. 2 Case #3: Cerebellary d e

pilocytic astrocytoma.
Preoperative (a), intraoperative
(b), and 5 year postoperative
image following resection (c).
Surgical position inside ‘‘double
doughnut’’ ioMRI (d, e). Posterior
fossa operations are made f Case #4 g h i
significantly more challenging by
requiring placement of the patient
superiorly and rotation of the
surgeon inferiorly (d). Case #4:
Midbrain cystic astrocytoma.
Note close proximity to midbrain
structures. Preop (f), intraop (g, h)
and 5 year postop images (i)
ioMRI Utilization in Cyst Management abscesses. When obtaining culture material we elected to
and CSF Diversion use the ioMRI facility due to the small size of these abscess-
es. With each catheter pass we could feel the tip hit the lesion
Time has also shown ioMRI to be particularly effective in and then slide off. With confirmation from intraoperative
surgeries requiring precise localization. In our pediatric neu- imaging the accuracy of our trajectory was not in question.
rosurgical experience this has been most appreciated in Therefore we felt confident in attempting more invasive
difficult CSF diversions, complex cysts drainage, and place- measures to perforate the abscesses wall. We enlarged the
ment of catheters near areas of eloquent anatomy. While burr hole slightly, followed the tract with bipolar cautery and
most static frameless stereotactic guidance systems which a #5 sucker, and entered the tense abscess rim exposing
rely on preoperative images can aid in surgical planning, purulent material. Cultures obtained during this procedure
these systems can not verify proper placement intraopera- grew streptococcus intermedius. Six month follow up imag-
tively. Additionally static stereotaxy does not offer a dynamic ing showed almost no sign of pathology.
view of operative changes (i.e. hemorrhage [14]) or guid- As stated before, ioMRI is most useful when its real-time
ance during the actual placement of the catheter because images drive surgical decision making. This is a function
they lack real-time image revision. In contrast, ioMRI offers that cannot be duplicated by static frameless stereotaxis
all three. An example of ioMRI driving surgical decision which relies on preoperative images. This is no truer than
making in shunt placement is Case #5, a 3 year old who while shunting loculated cysts. Case #7 is a 9 year old male
presented with pseudotumor cerebri, small ventricles, and who presented with signs and symptoms of increased intra-
intracranial hypertension (Fig. 3a d). Using ioMRI, we eas- cranial pressure. Subsequent imaging showed a complex
ily and precisely place a catheter into the ventricles on the cyst in the posterior right lateral ventricle (Fig. 3h k). The
first pass. In addition to CSF return, accurate placement of septa can be clearly appreciated on the preoperative images
the shunt was confirmed in the operating room prior to (Fig. 3h). On the initial catheter placement CSF was
closure. Within this ioMRI system spatial coordinate experi- obtained. However, after reviewing the intraoperative
ments reveal a mean overall error in acquisition of only images, it was clear that the catheter had deviated medially,
0.2 mm [14 16]. deflecting off the cysts (Fig. 3i). During a second attempt the
The accuracy of ioMRI is similarly illustrated by cases shunt can be seen piercing the septa as it traverses the cyst
such as that pictured in Fig. 3e g. Case #6 is a 13 year old and enters the left lateral ventricle, which was confirmed in
female with a seizure disorder who was admitted for increas- three plains (Fig. 3j). ioMRI, while substantially more ex-
ing seizure activity, right-sided weakness, and pneumonia. pensive that endoscopy, holds several key advantaged for
She was subsequently discovered to have multiple brain catheter placement as illustrated by this case. First, the
a b c d
Case #5
Fig. 3 Case #5: Intracranial

Hypertension with Slit Ventricles.
Preoperative CT (a),
intraoperative MRI showing stent e f g
Case #6
placement on first pass (b, c), and
post operative CT images (d).
Case #6: Abscess Drainage.
Preoperative image (e) showing
patient with multiple small strep
positive abscesses. Intraoperative
imaging immediately prior to and
following abscess drainage (f, g).
Case #7: Multiloculated h Case #7 i j k
Intraventricular Cyst.
Preoperative (h) and
intraoperative images showing a
missed first pass (i) and a
successful fenestration (j).
Postoperative image showing
decreased left lateral ventricle
The Evolution of ioMRI Utilization for Pediatric Neurosurgery: A Single Center Experience 93
altered anatomy of a complex cyst can be confusing. This ioMRI cases in our institution has actually decreased over
often results in the surgeon having to pierce tissue plains the last 10 years due to more selective use of this technology.
blindly without knowledge of which cavity or structures they However, while a majority of pediatric neurosurgical cases
might be entering. Secondly, endoscopic views are often can be performed in standard operating room; select cases,
obscured by blood or debris rendering the endoscope such as those detailed above, are made distinctly easier and
useless. Third, as with shunt placement, endoscopy only safer by utilizing ioMRI. This suggests that certain cases will
provides feedback once a cavity has been pierced while remain within the purview of this technology. Also given the
ioMRI allows guidance through parenchyma as well as ever expanding capabilities of MR the abilities of ioMRI
within fluid filled spaces. Fourth, ioMRI allows verification will only continue to expand [19].
of shunt placement and allows assurance that all loculi have
been violated prior to closure. Conflicts of Interest Statement We declare that we have no con
flict of interest.
In 1996 the few ioMRI systems under development were
predominantly housed within research universities. In 2009
over 100 systems are now in operation with a significant
number in large clinical centers, thus indicating a shift in
ioMRI from predominantly research applications to more References
widespread clinical usage. Additionally, while only 3% of
United States hospitals have access to ioMRI facilities, 17 of 1. Black PM, Moriarty T, Alexander E, Stieg P, Woodard EJ,
Gleason PL, Martin CH, Kikinis R, Schwartz RB, Jolesz FA
the 170 children’s hospitals have access, further indicating
(1997) Development and implementation of intraoperative
ioMRI’s usefulness in pediatric neurosurgical cases. Addi- magnetic resonance imaging and its neurosurgical applications.
tionally, private conversations within the field indicate Neurosurgery 41:831 842, Discussion 842 845
that many other pediatric institutions plan to acquire this 2. Tronnier VM, Wirtz CR, Knauth M, Lenz G, Pastyr O, Bonsanto
MM, Albert FK, Kuth R, Staubert A, Schlegel W, Sartor K, Kunze S
technology soon.
(1997) Intraoperative diagnostic and interventional magnetic
resonance imaging in neurosurgery. Neurosurgery 40:891 900,
Discussion 900 902
3. Schulder M, Jacobs A, Carmel PW (2001) Intraoperative MRI and
Summary adjuvant radiosurgery. Stereotact Funct Neurosurg 76:151 158
4. Hall WA, Martin AJ, Liu H, Pozza CH, Casey SO, Michel E,
Nussbaum ES, Maxwell RE, Truwit CL (1998) High field strength
The general advantages of ioMRI have been explained else- interventional magnetic resonance imaging for pediatric neurosur
where [15, 17, 18]. In brief the major advantages offered are gery. Pediatr Neurosurg 29:253 259
5. Sutherland GR, Louw DF (1999) Intraoperative MRI: a moving
continually updated dynamic images with a high level of magnet. CMAJ 161:1293
accuracy in a timely manner [14]. However, here we present 6. Mutchnick IS, Moriarty TM (2005) Neurosurgical uses for intra
lessons learned in 9 years of ioMRI utilization in relation to procedural magnetic resonance imaging. Top Magn Reson Imag
pediatric neurosurgery. In short, within this population ing 16:383 395
7. Shaw EG, Berkey B, Coons SW, Bullard D, Brachman D, Buckner
ioMRI is ideal for . . . JC, Stelzer KJ, Barger GR, Brown PD, Gilbert MR, Mehta M
1. Small, discrete tumors with poor normal/abnormal differ- (2008) Recurrence following neurosurgeon determined gross
total resection of adult supratentorial low grade glioma: results of
entiation. a prospective clinical trial. J Neurosurg 109:835 841
2. Large tumors with surrounding edema that . . . 8. Talos IF, Zou KH, Ohno Machado L, Bhagwat JG, Kikinis R,
Black PM, Jolesz FA (2006) Supratentorial low grade glioma
(a)Obscures the normal/abnormal differentiation, or resectability: statistical predictive analysis based on anatomic
(b)Causes a mass effect that will alter the anatomy sig- MR features and tumor characteristics. Radiology 239:506 513
nificantly from preoperative images after craniotomy. 9. Claus EB, Horlacher A, Hsu L, Schwartz RB, Dello Iacono D,
Talos F, Jolesz FA, Black PM (2005) Survival rates in patients with
3. Catheter/CSF diversion when the target . . . low grade glioma after intraoperative magnetic resonance image
guidance. Cancer 103:1227 1233
(a)Is small 10. Maksoud YA, Hahn YS, Engelhard HH (2002) Intracranial epen
(b)Is surrounded by eloquent anatomy, or dymoma. Neurosurg Focus 13:e4
11. Robertson PL, Zeltzer PM, Boyett JM, Rorke LB, Allen JC,
(c)Verification of shunt placement and not merely CSF Geyer JR, Stanley P, Li H, Albright AL, McGuire Cullen P, Finlay
return is required (i.e. multiloculated or intraventricu- JL, Stevens KR Jr, Milstein JM, Packer RJ, Wisoff J (1998)
lar cysts). Survival and prognostic factors following radiation therapy and
chemotherapy for ependymomas in children: a report of the
Children’s Cancer Group. J Neurosurg 88:695 703
The only relative contraindication that we have discov- 12. Gunny RS, Hayward RD, Phipps KP, Harding BN, Saunders DE
ered in this particular system is the posterior fossa approach. (2005) Spontaneous regression of residual low grade cerebellar
As a result of these observations, the number of pediatric pilocytic astrocytomas in children. Pediatr Radiol 35:1086 1091
13. Schulder M, Sernas TJ, Carmel PW (2003) Cranial surgery and 16. Schenck JF, Jolesz FA, Roemer PB, Cline HE, Lorensen WE,
navigation with a compact intraoperative MRI system. Acta Neu Kikinis R, Silverman SG, Hardy CJ, Barber WD, Laskaris ET
rochir Suppl 85:79 86 et al (1995) Superconducting open configuration MR imaging sys
14. Moriarty TM, Quinones Hinojosa A, Larson PS, Alexander E 3rd, tem for image guided therapy. Radiology 195:805 814
Gleason PL, Schwartz RB, Jolesz FA, Black PM (2000) Frameless 17. Vitaz TW, Hushek S, Shields CB, Moriarty T (2003) Intraoperative
stereotactic neurosurgery using intraoperative magnetic resonance MRI for pediatric tumor management. Acta Neurochir Suppl
imaging: stereotactic brain biopsy. Neurosurgery 47:1138 1145, 85:73 78
discussion 1145 1146 18. Woodard EJ, Leon SP, Moriarty TM, Quinones A, Zamani AA,
15. Moriarty TM, Kikinis R, Jolesz FA, Black PM, Alexander E 3rd Jolesz FA (2001) Initial experience with intraoperative magnetic
(1996) Magnetic resonance imaging therapy. Intraoperative MR resonance imaging in spine surgery. Spine 26:410 417
imaging. Neurosurg Clin N Am 7:323 331 19. Jolesz FA (2005) Future perspectives for intraoperative MRI. Neu
rosurg Clin N Am 16:201 213
Intraoperative MRI - High Field Systems
Implementation and Preliminary Clinical Experience with the
Use of Ceiling Mounted Mobile High Field Intraoperative Magnetic
Resonance Imaging Between Two Operating Rooms
Michael R. Chicoine, Chris C. H. Lim, John A. Evans, Amit Singla, Gregory J. Zipfel, Keith M. Rich, Joshua L. Dowling,
Jeffrey R. Leonard, Matthew D. Smyth, Paul Santiago, Eric C. Leuthardt, David D. Limbrick, and Ralph G. Dacey
Abstract Objective: Intraoperative magnetic resonance the clinical and financial impact of this technology in the
imaging (ioMRI) provides immediate feedback and quality field of neurosurgery.
assurance enabling the neurosurgeon to improve the quality
of a range of neurosurgical procedures. Implementation of Keywords Brain tumors Gliomas High field intraoperative
ioMRI is a complex and costly process. We describe our magnetic resonance imaging ioMRI Neurosurgery Pitui-
preliminary 16 months experience with the integration of an tary adenomas
IMRIS movable ceiling mounted high field (1.5 T) ioMRI
setup with two operating rooms.
Methods: Aspects of implementation of our ioMRI and our
initial 16 months of clinical experience in 180 consecutive Introduction
patients were reviewed.
Results: The installation of a ceiling mounted movable
Improvements in the operating theatre have been advancing
ioMRI between two operating rooms was completed in
for centuries [1], but never at such a rate seen in recent years.
April 2008 at Barnes-Jewish Hospital in St. Louis. Experi-
Applications of technological advancements are perhaps
ence with 180 neurosurgical cases (M:F 100:80, age range
more apparent in neurosurgery than in any other surgical
1 79 years, 71 gliomas, 57 pituitary adenomas, 9 metastases,
specialty. Over the past decade advancements in frameless
11 other tumor cases, 4 Chiari decompressions, 6 epilepsy
stereotaxy and ioMRI have enhanced the neurosurgeon’s
resections and 22 other miscellaneous procedures) demon-
capability to navigate the complex anatomy of the central
strated that this device effectively provided high quality
nervous system, improve on safety and efficacy of neurosur-
real-time intraoperative imaging. In 74 of all 180 cases
gical procedures and optimize patient outcomes for many
(41%) and in 54% of glioma resections, the surgeon mod-
diseases affecting the brain [2 12].
ified the procedure based upon the ioMRI. Ninety-three
A variety of ioMRI devices (high/low field, fixed/mov-
percent of ioMRI glioma cases achieved gross/near total
able magnets) can be implemented in neurosurgical
resection compared to 65% of non ioMRI glioma cases in
operating rooms. IoMRIs have been shown to improve the
this time frame.
extent of resection especially in gliomas and pituitary ade-
Conclusion: A movable high field strength ioMRI can be
nomas and this translates into improved patient outcomes
safely integrated between two neurosurgical operating
and survivals [9, 13 18].
rooms. This strategy leads to modification of the surgical
The newest generation of high-field strength ioMRIs
procedure in a significant number of cases, particularly for
offers potential for improvements yet to be realized by the
glioma surgery. Long-term follow up is needed to evaluate
earlier generations of these devices. Through collaboration
between Washington University, Barnes-Jewish Hospital,
St. Louis Children’s Hospital and IMRIS (Winnipeg, Man-
itoba, Canada) installation was completed of one of the first
M.R. Chicoine (*), C.C.H. Lim, J.A. Evans, A. Singla, G.J. Zipfel, movable ceiling mounted high field strength ioMRI devices
K.M. Rich, J.L. Dowling, J.R. Leonard, M.D. Smyth, P. Santiago,
E.C. Leuthardt, D.D. Limbrick, and R.G. Dacey in the world that can be accessed and utilized from either of
Department of Neurological Surgery,Washington University School of the two operating rooms (Fig. 1a). Integration of ioMRI into
Medicine, Campus Box 8057, 660 S. Euclid Avenue, St. Louis, MO the operating room is a complex and costly process. We
63110, USA describe the process of integration of this ioMRI device
e mail: chicoinem@wudosis.wustl.edu
98 M.R. Chicoine et al.
Fig. 1 (a) Line drawing

depicting the IMRIS 2 room
ioMRI model. Patient entry bore
of the magnet is rotated to face
either operating rooms within the
storage bay prior to moving out
from storage to its final position.
Gauss lines depicted (50G line
(red) and 5G line (black)) when
the ioMRI is in final imaging
position (image courtesy of
IMRIS); (b) sterile ‘‘C arm’’
drape covering patient and
operating table with upper portion
of 8 channel ioMRI coil in place;
(c) opening doors to the magnet
bay in preparation for moving
ioMRI device to the patient for
imaging (note the ceiling
mounted tracts); (d) after 90 s
ioMRI device is in its final
position for acquisition of images
into our operating rooms and our initial 16 months experi- (Phillips and Hitachi open MRI, Medtronic Odin, BrainLab
ence with 180 cases. BrainSuite and the IMRIS ceiling mounted ioMRI) and to
see them in use. High-field magnets provided better image
resolution and more versatility for current and future imag-
Methods ing modalities, but with greater associated costs and more
complex implementation issues. The BrainSuite offered
A review was made of the financial, architectural, engineer- appealing features including integration of surgical planning
ing, and marketing aspects of planning for and implementa- technology and high-resolution imaging. The IMRIS ioMRI
tion of an IMRIS ceiling mounted movable ioMRI device device is ceiling mounted and movable, the operating table
situated between two multi-purpose neurosurgical operating and patient remain in a fixed position and the device moves
rooms. to the patient. This feature in our view enhances safety in
A prospective database has been established and main- terms of airway control, monitoring and head fixation. Also
tained with approval of the Washington University Human for our high volume center, two-room ioMRI access would
Studies Institutional Review Board to monitor outcomes for add flexibility in terms of scheduling and room utilization.
all patients undergoing ioMRI procedures. A retrospective Standard surgical instruments could be used and these fully
analysis of the database was performed to assess the clinical equipped operating rooms can be adapted to serve as multi-
attributes, the impact on the surgeon’s decision-making pro- purpose operating rooms for any surgical case when the
cess, workflow issues and the preliminary outcomes of the ioMRI is not in use.
180 cases performed in the initial 16 months of experience. The installation of an ioMRI was associated with com-
plex architectural, engineering and safety issues and this was
integrated into a major renovation project encompassing
Results over 60 operating rooms with an estimated budget of US
$100 million over a 2 year period. Alternative solutions
including installation into the department of radiology
MRI Installation and Integration were impractical with respect to workflow, efficiency and
safety.
Preparation for installation of an ioMRI included analysis of The integration of ioMRI into our current neurosurgical
the commercially available ioMRI devices and our team armamentarium was intended to improve the precision,
visited other facilities internationally to assess these devices efficacy and safety of neurosurgical procedures. Within the
Implementation and Preliminary Clinical Experience with the Use of Ceiling Mounted 99
current coding system, direct billing and reimbursement for 44 years), 71 were gliomas, 57 pituitary adenomas, 9 metas-
the addition of the ioMRI is rather limited. A shared diag- tases, 5 craniopharyngiomas, 2 clival chordomas, 2 meningio-
nostic and intraoperative model could help reimburse the mas, 3 hemangioblastomas, 1 medulloblastoma, 1 trigeminal
cost from diagnostic revenues generated but was not feasible schwannoma, 6 resections for epilepsy, 4 Chiari decompres-
in our setup. Economic reimbursement stems from the po- sions and 19 other procedures for cyst aspiration/resection,
tential marketing impact. As one of the first centers world- biopsy or catheter placement. On 23 occasions, both of the
wide to employ this latest state of the art technology, it operating rooms were used simultaneously for 46 ioMRI-guid-
uniquely strengthens our position to offer improved patient ed procedures. In 74 of 180 cases (41%) the surgeon modified
care at local, regional, national, and international level. the surgical procedure based upon the findings of ioMRI.
Upon completion of the project, rigorous training pro- Gliomas: For the 71 glioma patients (34 high/37 low
grammes and simulations were instituted for all personnel grade), tumor resections were performed in 68 cases (32
involved to improve safety and efficiency. Access to the high/36 low grade) and the remainder underwent two biop-
ioMRI area is restricted through ioMRI security protected sies and one cyst aspiration. Awake techniques were incor-
doors. As a fully functional 1.5 T MRI device with a large porated into eight of these resections (six high/two low
patient-entry bore (70 cm), the Siemens Espree MRI scanner grades). Thirty-seven of 68 resections (54%) had additional
integrated into our IMRIS ioMRI suite offers full MRI tumor resected after ioMRI (Fig. 2). For the subset of 36
capability in a wide range of clinical and research utilization. low-grade gliomas (World Health Organization, WHO I II),
Some modifications of the surgical environment were additional tumor was resected in 15 cases (42%). Twenty-
necessary with the ioMRI. The IMRIS operating table has two of the 32 (69%) high-grade glioma (WHO II IV) patients
the ability to elevate, rotate, flex and had allowed us to had further resections. Pathological analysis of post ioMRI
perform our operations (146 supine, 27 prone and 7 lateral resection materials revealed tumor in 23 of the 29 specimens
cases) in almost any surgical position. The 3-pin ioMRI head received (79%). Six specimens were negative for tumor and
holder is made to hold the lower half of the MRI coil and additional eight specimens were not analyzed.
attaches to the operating table in a novel fashion. It has Pituitary Adenoma: Fifty-seven resections were per-
required some minor adjustments in order to incorporate it formed (55 transphenoidal, 2 transcranial) in 56 patients.
into routine use. The ioMRI compatible anesthesia machine IoMRI revealed possible residual tumor in 37 cases (65%)
perhaps does not have all of the capabilities of a standard (Fig. 3). Nineteen cases were not re-explored as it was felt
anesthesia machine but has been more than adequate for unsafe. Thirteen of the 18 explored had further resection, 4
even complex surgical procedure so far. had no tumor visualized and 1 had tumor encasing the
We have also incorporated the Medtronic Stealth frame- carotid. Of the nine post ioMRI resection pathology speci-
less stereotaxy and endoscopy systems with large high defi- mens received, eight were positive for tumor.
nition flat panel wall and touch screen monitors with Metastases: For the nine resections of brain metastases
tracking cameras on ceiling mounted booms into our performed, ioMRI after initial tumor resection suggested
ioMRI rooms which facilitate work flow well. residual tumor in five cases (56%). Four cases were explored
When the ioMRI is required, appropriate patient draping and histology revealed tumor in two.
and safety checks are completed (Fig. 1b). All equipment, Chiari Malformations: IoMRI was completed after sub-
surgical instruments, and ferromagnetic objects are moved occipital craniotomy and C-1 laminectomy with dural band
outside the 5 Gauss line (delineated with colored floor release without durotomy in four cases. In all cases ioMRI
tiling). A floor space in excess of 850 sq ft for each operating (anatomical) did not convincingly demonstrate adequate
room enables us to move the equipment to a safe distance decompression and therefore duraplasties were performed.
without having to remove the equipment from the room. The Post duraplasty ioMRI showed improved decompression in
ioMRI device moves from storage to imaging position in all cases.
90 s (Fig. 1c, d). Aspiration/Biopsy: The ioMRI has been used in 19 cases
to guide and assess the adequacy of biopsies, catheter inser-
tions and cyst aspirations/resections but did not lead to any
modifications of the procedures.
Initial 16 Months Clinical Experience Epilepsy Resections: Six cases used ioMRI to assess the
extent of resection for an epileptogenic focus. IoMRI con-
Between April 2008 and August 2009, 187 ioMRI cases were firmed accurate localization of resection in all. Three of six
performed. Seven cases were excluded (four planning MRI cases lead to additional surgical resection.
only, one image distortion from ET tube, one large body Wound Infections and Safety: Due to the increased
habitus and one technical failure). Of the remaining 180 duration of surgery and the additional sterile draping and
patients (M:F 100:80, age range 1 79 years, median undraping procedures necessary to perform an ioMRI, one
Fig. 2 Sample glioma case. (a) Preoperative MRI (left) of right frontal anaplastic oligodendroglioma prior to surgery. (b) ioMRI (middle) after
initial resection demonstrated residual tumor at the posterior margin of the resection cavity which prompted further resection of tumor.
(c) Postoperative MRI (right) demonstrating resection of the residual tumor identified on ioMRI
Fig. 3 Sample pituitary adenoma case. (a) Preoperative MRI (left) of large non functioning pituitary macro adenoma with suprasellar extension.
(b) ioMRI (middle) demonstrated evidence of residual tumor on the left side which prompted further resection of tumor. There is evidence of the
autologous adipose graft in the right aspect of the sella. (c) Postoperative MRI (right) demonstrating resection of the residual tumor identified on
the ioMRI
might predict an increased risk of surgical infection. Thus far used in tandem, total OR time averaged 466 min for glioma
no surgical wound infections have been identified for the (n¼16) and 398 min for pituitary adenoma (n¼17), com-
180 procedures that have been completed to date. Using a pared to 485 min for glioma (n¼44) and 423 min for pitui-
rigid set of safety guidelines, instrument counting and other tary adenoma (n¼38) when one room was used.
checklist strategies, no projectile objects or other safety
issues have been identified thus far with the use of this
movable high field ioMRI device.
Time intervals and workflow: Preparation time for ioMRI Discussion
averaged 21 min (7 66 min). Scanning time averaged
41 min (15 104 min). Post ioMRI time averaged 16 min An IMRIS ioMRI suite consisting of two operating rooms on
(3 63 min) to the start of further resection or wound closure. either side of a movable high field strength 1.5 T magnet was
Total operating room (OR) times were calculated for the successfully installed in April 2008. Barnes-Jewish Hospital
two largest subgroups, namely glioma (resection and non- is one of the few facilities in the world to have this advanced
awake cases) and pituitary adenoma (transphenoidal) to intraoperative imaging capability with access from two
determine if delay occurs when ioMRI cases were carried adjacent operating rooms. The concept of ioMRI has been
out in two rooms simultaneously. When two rooms were available for over a decade, but recent advances make this
Implementation and Preliminary Clinical Experience with the Use of Ceiling Mounted 101
technique more versatile than ever before. Intraoperative technology in the neurosurgical operating room remains to
MRI offers immediate feedback to the surgeon so that be seen.
the extent of resection might be determined. Increasingly The two-room model serves us well. There was no signif-
evidence indicates that a more complete surgical resection is icant workflow delay noted when the two ioMRI rooms were
associated with increased survival for both low and high- used in tandem.
grade gliomas, and that the extent of resection can be
improved with the use of ioMRI [14, 16, 17, 19, 20, 24].
Our preliminary experience with 68 glioma resections Conclusion
indicates that the MRI device enabled the neurosurgeon to
improve the extent of resection in 54% of cases, including
A movable high field strength ioMRI can be integrated into
42% for low-grade gliomas and 69% of high-grade gliomas.
neurosurgical practice and provide real time high quality
Ninety-three percent (n¼71) of ioMRI cases achieved gross/
imaging to guide the surgical treatment of brain tumors
near total resection compared to 65% (n¼53) of non ioMRI
and other diseases. This technique provides immediate feed-
cases in this time frame based on MRI scans within 72 h
back to assess the adequacy of surgery. This strategy leads
post-op. Continued monitoring of these patients will provide
to modification of the surgical procedure in a significant
information on the impact of ioMRI upon survival.
number of cases, particularly glioma surgery in which
Maximal safe resection is typically the goal for pituitary
ioMRI may suggest need for additional resection in every
surgery. Most lesions are adequately assessed via the trans-
second case. Long-term follow up is needed to evaluate the
phenoidal route, but there are limitations to visualization
clinical and financial impact of this technology in the field of
because of the depth of the surgical approach and the limited
neurosurgery.
corridor that can be establish through the anterior face of the
sella. The endoscope circumvents some of these limitations Conflicts of interest statement Dr. Chicoine has received an unre
by providing a more panoramic view especially when stricted educational grant from IMRIS, Winnepeg, Canada for mainte
coupled with contemporary high definition/3D video equip- nance of a database.
ments. Nonetheless, unintentional incomplete resection of
pituitary adenomas through the transphenoidal approach
remains a potential issue. Previous reports have shown References
residual adenoma in 39 66% leading to additional resection
in 22 66% of cases [9, 13]. Our initial experience with 57 1. Liu CY, Apuzzo MLJ (2003) The genesis of neurosurgery and the
cases is consistent with prior reports in that ioMRI showed evolution of the neurosurgical operative environment: Part I
Prehistory to 2003. Neurosurgery 52:3 19
evidence of probable residual tumor in 37 cases (65%).
2. Adeolu AA, Sutherland GR (2006) Intraoperative magnetic reso
Eighteen of these were explored (32%) and 13 had further nance imaging and meningioma surgery. West Afr J Med 25
resections, 4 were not visible and 1 found encasing the (3):174 178
carotid artery. With the continued evolution of surgical 3. Black PM, Alexander E 3rd, Martin C, Moriarty T, Nabavi A,
Wong TZ, Schwartz RB, Jolesz F (1999) Craniotomy for tumor
techniques for pituitary tumors in recent years including
treatment in an intraoperative magnetic resonance imaging unit.
the application of surgical navigation and endonasal endos- Neurosurgery 45(3):423 431, discussion 431 433
copy, the precise role and benefit of ioMRI remains to be 4. Black PM, Moriarty T, Alexander E 3rd, Stieg P, Woodard EJ,
defined. Gleason PL, Martin CH, Kikinis R, Schwartz RB, Jolesz FA (1997)
Surgical resection of brain metastases is well established,
nance imaging and its neurosurgical applications. Neurosurgery 41
particularly for solitary lesions. Metastatic tumors typically (4):831 842, discussion 842 845
appear well encapsulated and readily resectable yet surgery 5. Buchfelder M, Fahlbusch R, Ganslandt O, Stefan H, Nimsky C
alone is associated with significant rates of local recurrence. (2002) Use of intraoperative magnetic resonance imaging in tailored
temporal lobe surgeries for epilepsy. Epilepsia 43(8):864 873
Because of this, post-operative radiation is typically recom-
6. Kelly JJ, Hader WJ, Myles ST, Sutherland GR (2005) Epilepsy
mended [21 23]. IoMRI in our small series showed suspi- surgery with intraoperative MRI at 1.5 T. Neurosurg Clin North
cious areas in five of the nine patients, yet histopathological Am 16(1):173 183
evidence of tumor is verified in only two of the four cases 7. Lee MW, De Salles AA, Frighetto L, Torres R, Behnke E,
Bronstein JM (2005) Deep brain stimulation in intraoperative
explored. The role of ioMRI for resection of brain metas-
MRI environment comparison of imaging techniques and elec
tases and the significance of the intraoperative MRIs needs trode fixation methods. Minim Invasive Neurosurg 48(1):1 6
further refinement. 8. Mittal S, Black PM (2006) Intraoperative magnetic resonance
The number of cases in the remainder of this review is too imaging in neurosurgery: the Brigham concept. Acta Neurochir
Suppl 98:77 86
few to draw hard conclusions. These cases do show that 9. Nimsky C, von Keller B, Ganslandt O, Fahlbusch R (2006) Intrao
ioMRI can be used effectively for a wide range of neurosur- perative high field MRI in transsphenoidal surgery of hormonally
gical procedures but the full potential of application of this inactive pituitary macroadenomas. Neurosurgery 59(1):105 114
10. Tirakotai W, Sure U, Benes L, Krischek B, Bien S, Bertalanffy H primary supratentorial glioblastoma multiforme a quantitative
(2003) Image guided transsylvian transinsular approach for insular radiological analysis. Neuroradiology 47(7):489 500
cavernomas. Neurosurgery 53:1299 1305 18. Vitaz TW, Hushek S, Shields CB (2003) Moriarty T. Intraoperative
11. Tyler D, Mandybur G (1999) Interventional MRI guided stereo MRI for pediatric tumor management. Acta Neurochir Suppl
tactic aspiration of acute subacute intracerebral hematomas. 85:73 78
Stereotact Funct Neurosurg 72(2 4):129 135 19. Hirschberg H, Samset E, Hol PK, Tillung T, Lote K (2005) Impact
12. Walker DG, Talos F, Bromfield EB, Black PM (2002) Intraopera of intraoperative MRI on the surgical results for high grade glio
tive magnetic resonance for the surgical treatment of lesions pro mas. Minim Invasive Neurosurg 48(2):77 84
ducing seizures. J Clin Neurosci 9(5):515 520 20. Stummer W, Reulen HJ, Meinel T, Pichlmeier U, Schumacher W,
13. Bohinski RJ, Warnick RE, Gaskill Shipley MF, Zuccarello M, Tonn JC, Rohde V, Oppel F, Turowski B, Woiciechowsky C, Franz
van Loveren HR, Kormos DW, Tew JM Jr (2001) Intraoperative K, Pietsch T (2008) ALA Glioma Study Group. Extent of resection
magnetic resonance imaging to determine the extent of resection of and survival in glioblastoma multiforme: identification of and
pituitary macroadenomas during transsphenoidal microsurgery. adjustment for bias. Neurosurgery 62(3):564 576
Neurosurgery 49(5):1133 1143, discussion 1143 1144 21. Limbrick DD, Lusis EA, Chicoine MR, Rich KM, Dacey RG,
14. Claus EB, Horlacher A, Hsu L, Schwartz RB, Dello Iacono D, Dowling JL, Grubb RL, Filiput EA, Dryzmala RE, Mansur DB,
Talos F, Jolesz FA, Black PM (2005) Survival rates in patients with Malyapa R, Simpson JR (2009) Combined surgical resection and
low grade glioma after intraoperative magnetic resonance image stereotactic radiosurgery for treatment of cerebral metastases. Surg
guidance. Cancer 103(6):1227 1233 Neurol 71(3):280 288, discussion 288 289
15. Martin CH, Schwartz R, Jolesz F, Black PM (1999) Transsphenoi 22. Patchell RA, Tibbs PA, Regine WF et al (1998) Postoperative
dal resection of pituitary adenomas in an intraoperative MRI unit. radiotherapy in the treatment of single metastases to the brain: a
Pituitary 2(2):155 162 randomized trial. JAMA 280(17):1485 1489
16. Nimsky C, Ganslandt O, von Keller B, Fahlbusch R (2003) 23. Patchell RA, Tibbs PA, Walsh JW et al (1990) A randomized trial
Preliminary experience in glioma surgery with intraoperative of surgery in the treatment of single metastases to the brain. N Engl
high field MRI. Acta Neurochir Suppl 88:21 29 J Med 322(8):494 500
17. Schneider JP, Trantakis C, Rubach M, Schulz T, Dietrich J, 24. Oh DS, Black PM (2005) A low field intraoperative MRI system
Winkler D, Renner C, Schober R, Geiger K, Brosteanu O, Zimmer for glioma surgery: is it worthwhile? Neurosurg Clin North Am 16
C, Kahn T (2005) Intraoperative MRI to guide the resection of (1):135 141
High-Field ioMRI in Glioblastoma Surgery: Improvement
of Resection Radicality and Survival for the Patient?
H. Maximilian Mehdorn, Felix Schwartz, Stefan Dawirs, Jürgen Hedderich, Lutz Dörner, and Arya Nabavi
Abstract Since the first patients underwent intracranial Introduction

tumor removal with the radicality control of intraopera-
tive MRI (ioMRI) in September 2005 in our department, The value of aggressive cytoreduction in gliomas has been
the majority of operations performed in the ioMRI room discussed and established in recent years [1 12], yet, the use
have been indicated for high grade gliomas. In order to eluci- of intraoperative magnetic resonance imaging (ioMRI) has
date the role of ioMRI scanning in patients harboring high- been questioned in the context of high-grade glioma WHO
grade gliomas (HGG) on their survival, one hundred ninety grades III and IV (HGG) surgery. Several publications have
three patients with gliomas WHO grades III and IV were stressed the value of ioMRI in low grade gliomas (LGG)
operated either in a standard microsurgical neuronavigated [13 15], but only one group has published long-term results
fashion or using additionally ioMRI and were included in [13]. Therefore we undertook to collect and follow our glio-
a follow-up study. The series started with surgeries from ma patients series from the start of our clinical use of a 1.5 T
September 2005 until October 2007. Patient attribution to Philips Intera MRI machine [16, 17], short bore, integrated
the two groups was based on the logistical availability of into a new operating room fully equipped with microsurgical
the ioMRI on a scheduled surgery day, and on the assumed and neuronavigation (BrainLAB) facilities adjacent to other
‘‘difficulty’’ of the surgery based on the location of the glioma neurosurgical operating rooms with identical microsurgical
in or near to an eloquent area. Surgery was intended to be as and neuronavigation (BrainLAB) facilities [18]. ioMRI
radical as possible without reduction of quality of life. should add to the benefit of neuronavigation [19].
First surgery was performed in 103 patients (75 WHO IV Since first surgical brain tumor resection using intrao-
and 28 WHO III) and will be the main topic of this paper. In 60 perative MRI (ioMRI) control in September 2005 an addi-
patients, ioMRI was used, while in 43 patients standard micro- tional 425 procedures have been performed in the Dept. of
surgical neuronavigated resection techniques were applied. Neurosurgery at Kiel University Medical Center until late
Patients were followed in regular intervals mostly until May 2009. The majority of patients (N¼282) being treated
death. Statistical analysis showed a median survival time for for gliomas, this report will focus on the benefit and discuss
patients in whom ioMRI had been used of 20, 37 months some pitfalls encountered in this type of surgery.
compared to 10, 3 months in the cohort who had undergone
conventional microsurgical removal.
Major influencing concomitants were WHO grades and
age which were balanced in both groups. Material and Methods
Keywords High-grade glioma Intraoperative magnetic Out of 426 operations performed in the period of September
resonance imaging Karnofsky performance score 2005 until late May 2009, 76% of the operations were
Microsurgery Survival performed for gliomas, seventeen percent for transsphenoi-
dal surgery including pituitary adenomas, craniopharyngeo-
mas and clival chordomas. In the glioma group, fifty percent
of the patients harbored a WHO grade IV glioma, 21%
H.M. Mehdorn (*), F. Schwartz, S. Dawirs, J. Hedderich, L. Dörner, presented with a WHO grade III glioma and only 9% with
and A. Nabavi
a grade II glioma.
Dept of Neurosurgery, University Clinics of Schleswig Holstein
Campus Kiel, Arnold Heller Street 41, 24105, Kiel, Germany The patients were taken to the operating room with the
e mail: mehdorn@nch.uni kiel.de head fixed into a MRI compatible Mayfield three-pin head
104 H.M. Mehdorn et al.
holder attached to the sliding and rotating operating table Table 1 Extent of resection of gliomas WHO grades III and IV
derived from a Philips Medical Systems angiography table. No ioMRI ioMRI
The intraoperative imaging protocol evolving over time Radical resection intention preop No 16 20
included (1) preop T2fast images MRI scans to check for Yes 27 40
optimal head positioning within the magnetic field and any Radical resection seen on postop MRI No 21 23
Yes 22 37
abnormalities that might interfere with the later intraoperative
scans e.g. MRI incompatible pins, undetected metals etc. (2)
follow up phone calls were made to the patients, their families
Next, the patient is brought to the operating position by
or physicians. Among patients undergoing first surgery, 28
approximately 30 rotation of the table, draped in a regular
patients harbored a WHO grade III glioma, and 75 patients
manner, and surgery is performed as usual using regular neuro-
presented with a WHO grade IV glioma. All patients under-
surgical instruments including neuronavigation and operating
went subsequent radiochemotherapy with temozolamide
microscope. Stepwise tumor removal is controlled by repeated
(TMZ). A multivariate analysis was performed concerning
switch from white light to fluorescent light incorporated in our
WHO grade, use +/ of ioMRI, radical resection of tumor as
microscopes, using ALA fluorescence [20] nearly routinely for
seen on 24 h postoperative MRI scan, awake craniotomy
HGG surgery. When no more fluorescence is seen the patient’s
(which was used in 42 patients with tumors in eloquent
operating field is draped in such a fashion that he can be
areas), carmustine implants and others as age and sex.
brought back into the scanning position. T2 W images are
obtained in three dimensions, and depending on the dignity
of the tumor these images may serve to re-reference for neuro-
navigation and further tumor resection, or T1 W images with- Results
out and with contrast medium as well as DWI are additionally
obtained. Depending on the progress and difficulty of surgery, Comparing the extent of resection in combined grads III
additional intermediate studies are performed or a final scan is and IV gliomas, Table 1 shows that radical resection if
obtained prior to dural and wound closure. intended is more often achieved using ioMRI than not
In order to study the effect of ioMRI onto the clinical using it. This indicates that ioMRI has been able to guide
evolution of HGG patients, a study was performed analysing, and help the surgeon to a more radical surgical approach.
early in the clinical use of our ioMRI setting, patients Of course the question came up whether this more radical
who underwent surgery with ioMRI and comparing them to approach would result in a more frequent reduction of quali-
patients operated, in the same time, without ioMRI, using ty of life as assessed by Karnofsky performance score (KPS).
only regular microsurgical neuronavigated tumor removal. This was evaluated by comparing the KPS preoperatively to
In the time interval between September 2005 until October the KPS at 2 weeks postop. Figure 1 shows that there is
2007, a group of 193 patients were operated in our dept for indeed a reduction of KPS 2 weeks after surgery but for both
HGG. Follow up for these patients was continued until March groups of patients although slightly more prominent in the
2009. Patients are usually regularly seen as outpatients or ioMRI group.
100
50
–50 KPS praeOP

KPS postOP
Change of KPS
–100
no ioMRI
Fig. 1 Karnofsky performance ioMRI (p = 0.005)
score pre and 2 weeks postop (p = 0.033)
High-Field ioMRI in Glioblastoma Surgery: Improvement of Resection Radicality and Survival for the Patient? 105
Fig. 2 Survival curves all

primary high grade gliomas 1.0
ioMRI
(N 103)
no
yes
0.8 no (censored)
yes (censored)
0.6
survival
0.4
0.2
0.0
p = 0.034
0.00 12.00 24.00 36.00 48.00
total survival time (months)
Survival for both groups of patients was evaluated calcu- ‘‘simple’’ gliomas were operated in a standard fashion. This
lating Kaplan Meier survival curves. Figure 2 shows the also explains why we have so far used quite frequently, in 34
combined curves for all 103 HGG patients. It is evident patients, combined awake with ioMRI in order to prevent
that using ioMRI improves survival statistically (p¼0.034) additional neurological deficits to occur [21, 22].
while (not shown) survival is still slightly better even in So far in the series, no patient has suffered from adverse
WHO grade IV patients operated using ioMRI but no more events when using ioMRI, although some have complained
statistically different from patients operated without ioMRI. discomfort when undergoing awake craniotomy and ioMRI
[23]. On-the-spot interpretation of ioMRI images by the
surgeon requires a certain knowledge related to contrast
Discussion media dynamics [24], particularly when performing repeat
T1W examinations. Some help is provided by computer
programs which our group has evaluated on a larger scale
The data presented here show that by adding ioMRI into the and uses now frequently to detect minor tumor remnants [25].
armamentarium of modern neurosurgery it is possible to im- The benefit of answering the question of brain shift by
prove overall survival measured in median survival time even using ioMRI is evident [26]. The socio-economic question,
for HGG patients. Due to the protocol implied in this group of however, of using ioMRI cannot be addressed here since it
patients using if possible both ALA-fluorescence and ioMRI it touches the treatment of malignant tumors generally spoken.
is formally impossible to differentiate the benefit of each of the In the same sense, every dept. has to carefully decide whether
two adjuncts to radical glioma surgery. This would obviously it can take the burden of implementing a high-field ioMRI
be interesting but rather on an academic or economical point of into its budget [27, 28]. Other options besides dedicated
view. Further studies are under way in this regard. neurosurgical ioMRI machines are certainly possible. On the
In the longterm follow up after 32 months, the survival other hand, the possibility to work around such a dedicated
curves approach each other. This confirms data obtained e.g. machine offers additional research incentives () to better
in the ALA studies [20] and points to the fact that additional understand the brain, its anatomy and treatment options.
therapy is required to further improve the longterm survival
in these patients. Conflicts of Interest Statement We declare that we have no
The data are important since they are derived from an conflict of interest.
unselected cohort of patients from a primary university referral
center favoring aggressive surgery even in patients when other
centers might have adopted a rather conservative approach. References
Attribution to one of the two treatment groups was performed
rather on a logistical than on a randomized basis: this means 1. Durmaz R, Erken S, Arslantaş A, Atasoy MA, Bal C, Tel E (1997)
that patients thought to harbor a ‘‘difficult’’ HGG in or close Management of glioblastoma multiforme: with special reference to
to an eloquent area underwent surgery using ioMRI while recurrence. Clin Neurol Neurosurg 99(2):117 123
106 H.M. Mehdorn et al.
2. Hall WA (2004) Extending survival in gliomas: surgical resection 16. Hall WA, Martin AJ, Liu H, Pozza CH, Casey SO, Michel E,
or immunotherapy? Surg Neurol 61(2):145 148 Nussbaum ES, Maxwell RE, Truwit CL (1998) High field strength
3. Jeremic B, Grujicic D, Antunovic V, Djuric L, Stojanovic M, interventional magnetic resonance imaging for pediatric neuro
Shibamoto Y (1994) Influence of extent of surgery and tumor surgery. Pediatr Neurosurg 29(5):253 259
location on treatment outcome of patients with glioblastoma multi 17. Martin AJ, Hall WA, Liu H, Pozza CH, Michel E, Casey SO,
forme treated with combined modality approach. J Neurooncol 21 Maxwell RE, Truwit CL (2000) Brain tumor resection: intra
(2):177 185 operative monitoring with high field strength MR imaging initial
4. Knauth M, Wirtz CR, Tronnier VM, Aras N, Kunze S, Sartor K results. Radiology 215(1):221 228
(1999) Intraoperative MR imaging increases the extent of tumor 18. Nabavi A, Dörner L, Stark AM, Mehdorn HM (2009) Intraopera
resection in patients with high grade gliomas. Am J Neuroradiol tive MRI with 1.5 Tesla in neurosurgery. Neurosurg Clin North Am
20(9):1642 1646 20(2):163 171
5. Laws ER, Parney IF, Huang W, Anderson F, Morris AM, Asher A, 19. Wirtz CR, Albert FK, Schwaderer M, Heuer C, Staubert A,
Lillehei KO, Bernstein M, Brem H, Sloan A, Berger MS, Chang S, Tronnier VM, Knauth M, Kunze S (2000) The benefit of neurona
Investigators GO (2003) Survival following surgery and prognostic vigation for neurosurgery analyzed by its impact on glioblastoma
factors for recently diagnosed malignant glioma: data from the surgery. Neurol Res 22(4):354 360
Glioma Outcomes Project. J Neurosurg 99(3):467 473 20. Stummer W, Pichlmeier U, Meinel T, Wiestler OD, Zanella F,
6. Rostomily RC, Spence AM, Duong D, McCormick K, Bland M, Reulen HJ, ALA Glioma Study Group (2006) Fluorescence
Berger MS (1994) Multimodality management of recurrent adult guided surgery with 5 aminolevulinic acid for resection of malig
malignant gliomas: results of a phase II multiagent chemotherapy nant glioma: a randomised controlled multicentre phase III trial.
study and analysis of cytoreductive surgery. Neurosurgery 35 Lancet Oncol 27(5):392 401
(3):378 388 21. Nabavi A, Goebel S, Doerner L, Warneke N, Ulmer S, Mehdorn M
7. Salcman M (1988) Surgical resection of malignant brain tumors: (2009) Awake craniotomy and intraoperative magnetic resonance
who benefits? Oncology 2(8):47 56 imaging: patient selection, preparation, and technique. Top Magn
8. Shinoda J, Sakai N, Murase S, Yano H, Matsuhisa T, Funakoshi T Reson Imaging 19(4):191 196
(2001) Selection of eligible patients with supratentorial glioblastoma 22. Pinsker MO, Nabavi A, Mehdorn HM (2007) Neuronavigation
multiforme for gross total resection. J Neurooncol 52(2):161 171 and resection of lesions located in eloquent brain areas under
9. Stark AM, Nabavi A, Mehdorn HM, Blömer U (2005) Glioblasto local anesthesia and neuropsychological neurophysiological moni
ma multiforme report of 267 cases treated at a single institution. toring. Minim Invasive Neurosurg 50(5):281 284
Surg Neurol 63(2):162 169 23. Goebel S, Nabavi A, Schubert S, Mehdorn HM (2010) Patient
10. Stark AM, Hedderich J, Held Feindt J, Mehdorn HM (2007) perception of combined awake brain tumour surgery and intra
Glioblastoma the consequences of advanced patient age on treat operative 1.5 T MRI: The Kiel experience. Neurosurgery
ment and survival. Neurosurg Rev 30(1):56 66 24. Knauth M, Wirtz CR, Aras N, Sartor K (2001) Low field interven
11. Vuorinen V, Hinkka S, Färkkilä M, Jääskeläinen J (2003) Debulk tional MRI in neurosurgery: finding the right dose of contrast
ing or biopsy of malignant glioma in elderly people a randomised medium. Neuroradiology 43(3):254 258
study. Acta Neurochir 145(1):5 10 25. Ulmer S, Helle M, Jansen O, Mehdorn HM, Nabavi A (2009)
12. Yoshida J, Kajita Y, Wakabayashi T, Sugita K (1994) Long term Intraoperative dynamic susceptibility contrast weighted magnetic
follow up results of 175 patients with malignant glioma: importance resonance imaging (iDSC MRI) Technical considerations and
of radical tumour resection and postoperative adjuvant therapy with feasibility. Neuroimage 45(1):38 43
interferon, ACNU and radiation. Acta Neurochir 127(1 2):55 59 26. Nimsky C, Ganslandt O, Cerny S, Hastreiter P, Greiner G,
13. Claus EB, Horlacher A, Hsu L, Schwartz RB, Dello Iacono D, Fahlbusch R (2000) Quantification of, visualization of, and com
Talos F, Jolesz FA, Black PM (2005) Survival rates in patients with pensation for brain shift using intraoperative magnetic resonance
low grade glioma after intraoperative magnetic resonance image imaging. Neurosurgery 47(5):1070 1079
guidance. Cancer 103(6):1227 1233 27. Oh DS, Black PM (2005) A low field intraoperative MRI system
14. Fahlbusch R, Ganslandt O, Nimsky C (2000) Intraoperative ima for glioma surgery: is it worthwhile? Neurosurg Clin N Am 16(1):
ging with open magnetic resonance imaging and neuronavigation. 135 141, Review
Childs Nerv Syst 16(10 11):829 831 28. Schneider JP, Trantakis C, Schulz T, Dietrich J, Kahn T
15. Nimsky C, Ganslandt O, Von Keller B, Romstöck J, Fahlbusch R (2003) [Intraoperative use of an open mid field MR scanner
(2004) Intraoperative high field strength MR imaging: implemen in the surgical treatment of cerebral gliomas]. Z Med Phys 13(3):
tation and experience in 200 patients. Radiology 233(1):67 78 214 218
Image Guided Aneurysm Surgery in a Brainsuite1 ioMRI Miyabi 1.5 T
Environment
R.W. König, C.P.G. Heinen, G. Antoniadis, T. Kapapa, M.T. Pedro, A. Gardill, C.R. Wirtz, T. Kretschmer, and T. Schmidt
Abstract guided aneurysm surgery in an ioMRI-environment may

Objective: Current literature only gives sparse account of be helpful especially in complex aneurysms and provide
aneurysm surgery in an intraoperative MRI environment. neurosurgeons and neuroanaesthesiologists with additional
After installation of a BrainSuite1 ioMRI Miyabi 1.5 T at information about cerebral haemodynamics and perfusion
our institution the aim of the present preliminary study was pattern in the vascular territory distal to the target vessel.
to evaluate feasibility, pros and cons of aneurysm surgery in
this special setting. Keywords Aneurysm clipping BrainSuite Image guidance
Material and Methods: Since February 2009, during a ioMRI Intraoperative imaging
3 months period we performed elective image guided aneu-
rysm surgery in 4 ACM and 1 ACOM aneurysm (four
patients) in this ioMRI setting. The patients’ heads were
Introduction
rigidly fixed in the Noras 8-Channel OR Head Coil. Our
imaging protocol included MP-RAGE, T2-TSE axial, TOF- After its introduction intraoperative MRI (ioMRI) rapidly
MRA and diffusion-/perfusion-imaging immediately before became the imaging modality of choice for central nervous
surgery and after clip application. Presurgical 3D-planning system pathology, particularly with regard to resection con-
was performed using the iPlan-Software. trol in glioma and pituitary surgery [1 5]. Additionally
Results: All five aneurysms were operated without tem- numerous articles published till now discuss various aspects
porary clipping. There were no intra- or postoperative com- of image guidance in aneurysm surgery [6 9].
plications. Patient positioning and head fixation with the Sutherland [10] was the first who reported about one case
integrated Noras Head Clamp was feasible, but there were of elective clipping of an ACOM aneurysm in an ioMRI
significant limitations particularly with regard to more com- environment especially accentuating the significance of
plex approaches and patient physiognomy. Image quality MRA and diffusion weighted imaging.
especially TOF-MRA was good in 4, insufficient in 1 aneu- This is the first study reporting about elective aneurysm
rysm. Presurgical planning especially vessel extraction surgery in four patients in a fully integrated neurosurgical
from TOF-MRA was possible but certainly needs significant OR (BrainSUITE1 ioMRI Miyabi 1.5 T; BrainLAB, Ger-
future improvement. Diffusion- and perfusion weighted many) combining image guidance with intraoperative im-
examinations yielded good image quality. aging particularly diffusion and perfusion weighted
Conclusion: Our limited experience is encouraging so far. imaging. The aim of the present study was to highlight
Further improvement particularly concerning flexibility of the aspects of feasibility and diagnostic relevance of MR-
patient positioning and presurgical 3D-planning for vascular TOF-Angiography and Perfusion Weighted Imaging (PWI)
procedures is most necessary. As a future perspective image before and after clip application in our special setting.
R.W. König (*), C.P.G. Heinen, G. Antoniadis, T. Kapapa,

M.T. Pedro, A. Gardill, C.R. Wirtz, and T. Schmidt
Department of Neurosurgery, University of Ulm, Ludwig Heilmeyer Str.
2, 89312, BKH, Guenzburg, Germany
The fully integrated single-room BrainSUITE1 ioMRI
e mail: ralph.koenig@uni ulm.de
T. Kretschmer Miyabi consists of a 1.5 T Siemens MAGNETOM Espree
Department of Neurosurgery, Ev. Krankenhaus, Oldenburg, Germany scanner combined with the Miyabi table concept. Hereby the
108 R.W. König et al.
Fig. 1 Noras 8 Channel OR

Head Coil with integrated head
holder (with permission of Noras
MRI products GmbH, Germany)
patient is positioned on the specially designed Miyabi shell, time 105 min) due to scanning before surgery and after clip-
which is mounted to a regular Trumpf Jupiter OR table ping and because of some limitations in patient positioning.
(TRUMPF Medizinsysteme, Germany). The layout chosen Infections as possible consequence of prolonged surgical
allows the neurosurgeon to use conventional surgical instru- procedures did not occur.
ments beyond the 5-G line. The patient’s head is rigidly Feasibility: All patients were operated via a standard
fixed in a head holder integrated in the Noras 8-Channel frontolateral approach. As a consequence patient positioning
OR Head Coil (NORAS MRI products GmbH, Germany). In was feasible but needed some additional preparation time
contrast to the BrainSUITE1 ioMRI rotating table a height before surgery. Flexibility of head fixation (tilting, vertex
adjustable fixation of the head coil is not possible, the coil is down positioning) is markedly restricted primarily due to the
mounted to the Miyabi shell with a semi-spherical joint fixation of the head coil with the Miyabi shell, which allows
(Fig. 1). only a limited range of movement. Due to missing height
Since February 2009, during a 3 months period we per- adjustment in the fixation of the head coil neurosurgical
formed elective image guided aneurysm surgery in 4 ACM procedures in prone position (so far no experience in aneu-
and 1 ACOM aneurysm (four patients). The imaging proto- rysm surgery, only in AVM surgery) are feasible but need
col included MP-RAGE, T2-TSE axial, TOF-MRA and dif- elaborate padding under the breast of the patient. Obese
fusion-/perfusion-imaging immediately before surgery and patients and those with unfavourable physiognomy (short
after clip application. Presurgical planning was performed neck, broad shoulders) should not be considered as surgical
using the iPlan1-Software (BrainLAB AG, Germany). candidates in the ioMRI environment described.
Results Diagnostic relevance of MR-TOF-Angiography

Before and After Clip Application
Patient Outcome: Management of all five aneurysms was
uneventful without any temporary clipping. As expected the We used MR-TOF-Angiography as basis of image guided
overall procedure time was extended (average additional aneurysm surgery (Fig. 2). The image quality was excellent
Image Guided Aneurysm Surgery in a Brainsuite1 ioMRI Miyabi 1.5 T Environment 109
Fig. 2 Image guided surgery

of left sided MCA bifurcation
aneurysm
Fig. 3 TOF MRA before and

after clipping of a left sided MCA
bifurcation aneurysm
in 3 MCA and one ACOM aneurysms correlating very well not be possible without significant reduction of these arte-
with preoperative 3D-CT-angiographic studies (3D-CTA) facts. Patency of the efferent vessels distal to the clipping
and intraoperative findings. On the other hand TOF-MRA site could be demonstrated in all cases.
underestimated size and configuration of one left sided MCA
bifurcation aneurysm. The main aneurysm sac (1.2 cm) pro-
jected mainly downward and therefore out of the vector of the
main blood flow, therefore TOF-MRA demonstrated the Perfusion Weighted Imaging Before and After
inflow zone of the aneurysm at its neck but not the complete Clip Application
lesion.
Intraoperative MRA after clip application revealed sig-
nificant susceptibility artefacts around the clip (Fig. 3a, b). Image quality of PWI before and after clip application
Hence judgement of aneurysm remnants or clip stenosis will was good. Besides the clip-induced artefact PWI could
110 R.W. König et al.
demonstrate normal haemodynamics and perfusion of both Conflicts of interest statement We declare that we have no conflict
hemispheres, especially in the vascular territory distal to the of interest.
aneurysm. In one patient after uneventful clipping of a MCA
aneurysm there was a slight delay in time to peak and mean References
transit time in the territory of the posterior parietal artery.
Recovery of the patient was normal and follow-up studies
1. Black PM, Moriarty T, Aleander E III, Stieg P, Woodard EJ,
did not reveal any infarctions. Gleason PL, Martin CH, Kikinis R, Schwartz RB, Jolest FA
(1997) Development and implementation of intraoperative mag
netic resonance imaging and its neurosurgical applications. Neuro
surgery 41:831 842
2. Tronnier VM, Wirtz CR, Knauth M, Lenz G, Pastyr O, Bonsanto MM,
Discussion Albert FK, Kuth R, Staubert A, Schlegel W, Sartor K, Kunze S
(1997) Intraoperative diagnostic and interventional magnetic reso
nance imaging in neurosurgery. Neurosurgery 40:891 900
To our knowledge this is the first study reporting about 3. Nimsky C, Ganslandt O, Kober H, Buchfelder M, Fahlbusch R
elective aneurysm surgery in a fully integrated neurosurgical (2001) Intraoperative magnetic resonance imaging combined with
OR (BrainSUITE1 ioMRI Miyabi 1.5 T; BrainLAB, Ger- neuronavigation: a new concept. Neurosurgery 48:1082 1089
4. Bradley WG (2002) Achieving gross total resection of brain
many) combining image guidance with advanced intraopera-
tumors: intraoperative MR imaging can make a big difference.
tive imaging particularly diffusion and perfusion weighted Am J Neuroradiol 23:348 349
imaging. 5. Fahlbusch R, Ganslandt O, Buchfelder M, Schott W, Nimsky C
The pioneer study of Sutherland [10] already demon- (2001) Intraoperative magnetic resonance imaging during trans
sphenoidal surgery. J Neurosurg 95:381 390
strated feasibility of elective aneurysm surgery in an
6. Albert FK, Wirtz CR, Forsting M, Jansen O, Polarz H, Mittermaier G,
ioMRI environment with a ceiling mounted moveable Kunze S (1998) Image guided excision of a ruptured feeding artery
1.5 T magnet. Image guidance was not integrated. The ‘‘pedicle aneurysm’’ associated with an arteriovenous malforma
study especially stressed the diagnostic relevance of MRA tion in a child. Case report. Comput Aided Surg 2:5 10
7. Schmid Elsaesser R, Muacevic A, Holtmannspötter M, Uhl E,
and intraoperative diffusion weighted imaging. Reduction of
Steiger HJ (2003) Neuronavigation based on CT angiography for
image distorsions caused by the aneurysm clip is one big surgery of intracranial aneurysms: primary experience with unrup
issue discussed there. We believe that further decrease of tured aneurysms. Minim Invasive Neurosurg 46:269 277
MR image artefacts aiming at judgement of probable aneu- 8. Pirotte B, Wikler D, David P, Lefranc F, Brotchi J, Levivier M
(2004) Magnetic resonance angiography image guidance for the
rysm remnants or vessel stenosis after clipping will be hard
microsurgical clipping of intracranial aneurysms: a report of two
to achieve. But this intraoperative diagnostic flaw can easily cases. Neurol Res 26:429 34
be compensated by adding e.g. microscope integrated near 9. Raabe A, Beck J, Rohde S, Berkefeld J, Seifert V (2006) Three
infrared indocyanine green video angiography [11]. dimensional rotational angiography guidance for aneurysm sur
gery. J Neurosurg 105:406 411
The combination of different advanced intraoperative
10. Sutherland G, Kalibara T, Wallace C, Boguslaw T, Richter M
imaging modalities like ioMRI (TOF, PWI, DWI, arterial (2002) Intraoperative assessment of aneurysm clipping using mag
spin labelling) and ICG-videoangiography might be helpful netic resonance angiography and diffusion weighted imaging:
in selected aneurysm cases. On the other hand there are some technical case report. Neurosurgery 50:893 898
11. Raabe A, Beck J, Gerlach R, Zimmermann M, Seifert V (2003)
disadvantages especially concerning patient positioning
Near infrared indocyanine green video angiography: a new method
which have to be balanced with the potential advantages for intraoperative assessment of vascular flow. Neurosurgery
achieved by performing aneurysm surgery in an ioMRI 52:132 139
environment.
From Intraoperative Angiography to Advanced Intraoperative
Imaging: The Geneva Experience
Karl Schaller, Marc Kotowski, Vitor Pereira, Daniel Rüfenacht, and Philippe Bijlenga
Abstract Objective: We aimed at the integration of recent Keywords Flat panel Intraoperative angiography Intrao-
flat panel technology in a joint interventional suite for neu- perative CT Intraoperative imaging Vascular neurosur-
rosurgeons and neuroradiologists. gery
Methods: A Flat Panel system, allowing for intraopera-
tive performance of 2D and 3D DSA, for automated seg-
mentation of vascular structures, and for performance of
computed tomography, was connected with a surgical mi-
croscope and neuronavigation. All surgical and neurointer- Introduction
ventional cases were monitored and stored in a prospective
data base. Neurosurgeons deal with intracranial and spinal vascular,
Results: N¼99 patients were treated neurosurgically: bony, and soft tissue pathologies, or with a combination of
N¼63 aneurysm clippings in n¼51 patients; n¼12 those. Due to its strength for the visualization of soft-tissue
resections of arteriovenous malformations (AVM); n ¼6 properties, intraoperative MRI (ioMRI) has become an
clippings/excisions of dural AV fistulae (dAVF); n¼3 important tool for resection control in surgery of intra-axial
EC-IC bypass procedures; n¼10 resections of skull brain tumours, and in pituitary surgery [1]. It requires large
base tumours; n¼17 spine procedures. All patients had and expensive equipment, however, and its use is still not
intraoperative imaging for angiographic control and/or for very intuitive in a strictly surgical sense. Intraoperative CT
anatomical allocation. Intraoperative 3D-rotational angi- (ioCT) scanning requires large equipment as well, and inter-
ography was performed n¼54 times in n¼42 patients in feres with the natural course of surgery as the patient needs
<15 min each, with repositioning of aneurysm clips in to be moved into the gantry [2]. Furthermore, vascular
n¼9 patients. imaging does not allow for dynamic visualization of blood
Conclusion: This hybrid neuro-interventional suite flow.
opens a new avenue for intraoperative imaging by the Recent Flat Panel (FP) technology allows for imaging of
provision of highly resoluted angiographic or CT images, vascular structures at high spatial and temporal resolution
which may be co-registered with a navigation system. In with up to 100 images/s. In addition, it allows for bone
addition, the workflow in treatment of aneurysmal SAH imaging at high spatial resolution, e.g. precise visualization
can be improved, as all diagnostic and therapeutic mea- of the human skull base, or of trabecular bone of the human
sures can be taken without having to move the patient to spine. The hardware itself is smaller than previously used
other facilities. X-ray machines, and it can be ceiling-mounted, or attached
to robotic arms. This technology may represent an interest-
ing alternative to the above-mentioned modalities, ioMRI
and ioCT due to its flexibility and due to the fact that it
allows for dynamic vascular imaging, as well as for flat
panel derived computed tomography as well. In addition,
adaptation of image-guidance could allow for co-registration
of the various image modalities taking advantage of each
K. Schaller (*), M. Kotowski, V. Pereira, D. Rüfenacht, and P. Bijlenga
of those. It was our ambition to develop and to evaluate a
University of Geneva Medical Center, Faculty of Medicine, University
of Geneva, Rue Gabrielle Perret Gentil 4, 1211 Geneva, Switzerland joint hybrid neuroradiological and neurosurgical interven-
e mail: karl.schaller@hcuge.ch tional suite, based on FP technology (Figs. 1 3).
112 K. Schaller et al.

Infrastructure
The hybrid interventional room is located between diagnos-

tic radiological facilities, such as MRIs, the emergency
surgical suites of the hospital, and the urological operating
rooms. It can be entered from three sides. The assigned
interventional room surface was restricted to 65 m2 for
local architectural reasons. In addition, there is a control
room with all monitors and computer workstations, shielded
by protecting glass towards the interventional suite, and a
small storage room with a scrubbing facility adding to
another 25 m2. Local bioengineers have evaluated the hygie-
nical aspects, as the room was intended for use in diagnostic
Fig. 1 Illustration of basic setup in the hybrid room. The Flat Panel and in open surgical interventions. Therefore, a vertical
system, including table with radiolucent headholder, and overhead moni laminar air flow system had to be installed at the foot end
tors. Not in the figure: Surgical microscope and neuronavigation, which of the interventional table.
are coupled with the FP system. 3D screen and Neuromonitoring system
Fig. 2 Example of joint

neurosurgical neuroradiological
intervention. Sixty eight year old
woman with incidentally found
aneurysms of her right superior
cerebellar (SCA) (a), anterior
communicating (Acom) (b), left
middle cerebral (MCA) (c), and
left posterior communicating
(Pcom) (c) arteries. All
aneurysms were treated the same
day, on the same table, during the
same session. Interventional
neuroradiologists started with coil
occlusion of the right SCA
aneurysm. These images
underscore the quality of the 3D
angiographic images
From Intraoperative Angiography to Advanced Intraoperative Imaging: The Geneva Experience 113
Fig. 3 Intra operative

angiographic control, showing
complete occlusion of all
aneurysms (2D DSA) in anterior
posterior (a) and lateral (b) view.
Please note the limited artefacts
of the headholder
Equipment patients, n¼12 resections of arteriovenous malformations

(AVMs), n¼6 clippings/excisions of dural AV fistulae
(dAVF), n¼3 EC-IC bypass procedures, n¼10 resections
A monoplane Philips Allura FD20 system (620 projections
of skull base tumours, and n¼17 complex spine procedures.
along 240 in 8 10 s, rotational speed: 30 /55 /s, 30 frames/s),
Intraoperative 3D-rotational angiography was performed
allowing for automated segmentation of vascular structures
in all vascular patients in <15 min each. It allowed for direct
was installed (Philips Medical systems, Best, Netherlands).
control of aneurysm-clipping and displayed incomplete clip-
This system allows for dynamic imaging (e.g. angiography),
ping (n¼7) or vascular compromise (n¼2). Direct reposition-
as well as for CT imaging. The accompanying (angio-)table
ing of aneurysm clips was performed in n¼9 patients.
can be adjusted in height only in its present version.
Thereby, intraoperative 3D angiography had stipulated clip
A special adapter for the radiolucent head clamp (Doro,
correction in 14.3% of aneurysms, thus allowing complete
Promedics, Düsseldorf, Germany) was constructed.
aneurysm occlusion in 100%. Intraoperative FP-based CT
A BrainLAB neuronavigation system (VV2; BrainLAB,
scanning was performed in skull base tumours. It was particu-
Heimstetten, Germany) was installed in addition, and a
larly helpful, where a far lateral transcondylar approach had to
recent surgical microscope (Zeiss Pentero; Zeiss, Oberkochen,
be used in a recurrent foramen magnum meningioma, and in a
Germany) including an Indocyanine green (ICG) videoangio-
partially calcified clival chordoma (see Figs. 4 and 5). The
graphy unit completed the basic equipment. The microscope
respective intraoperative data sets were co-registered with
is navigable, and the navigation system was adapted for
preoperative images on the navigation system. However, soft
integration and co-registration of FP-derived images (angio-
tissue properties could not be visualised with the same quali-
graphy and CT) with conventional imaging data sets.
ty as with extraoperative CT and certainly not as with MRI.
All surgical and neuroradiological and joint interventions
Two patients of the whole series developed locally con-
were prospectively stored in a data base.
fined decubitus following prolonged lateral positioning.
All patients had intra-operative imaging for anatomical
allocation (e.g. FP CT scanning for resection control in skull
base tumours) or for angiographic control (e.g. of clipped Neuroradiological Patients
aneurysms). Co-registration of intraoperatively acquired
imaging data (either angiographic, or FP CT) with the navi-
gation system was performed when considered useful for the Within the same period, a total of n¼147 diagnostic and
intervention. n¼106 interventional (neuroradiological) procedures were
performed (n¼51 coilings, n¼28 stentings, n¼13 AVM- or
AVF-embolizations, n¼14 vertebroplasties). In addition,
n¼364 percutaneous infiltrations were performed.
Results
Open Neurosurgical Patients Discussion
A total of n¼99 patients were treated neurosurgically within In the same way, neurosurgeons are eager to perform resec-
a 19-month period of time (February 2008 September tion control in case of intrinsic brain tumours by intra-
2009). This includes n¼63 aneurysm clippings in n¼51 operative MRI, they are looking for innovative ways of
114 K. Schaller et al.
Fig. 4 Intra operative 3D skull

radiography during extended far
lateral approach for a skull base
tumour. These images can be
transferred and co registered with
the navigation station
intraoperative visualization in neurovascular surgery, in sur- intra-operative CT scanners, the most recent of which do
gery around the skull base, and for spine surgery. That goal even allow for intraoperative performance of angio-CT scan-
should be accomplished in an intuitive manner, and what- ning in addition [9]. These scanners are large, however,
ever methodology applied it should not interfere too much fairly static, and it thus seems reasonable to combine the
with the surgical workflow. As the physical properties of the technical and physical flexibility of recent Flat Panel tech-
tissues neurosurgeons are dealing with are so different from nology in order to overcome the shortcomings of standard
each other, it is very unlikely to achieve this goal with a 2D DSA and CT scanning.
single imaging modality in the near future. Whereas the direct advantage of such a hybrid interven-
As far as it concerns vascular imaging, the dynamics of tional environment may not be as obvious for surgery of
intravascular flow have to be considered in addition. Where- intracranial tumours, it has to be stated that the integration
as ioMRI is steadily advancing, e.g. in the field of glioma of such a joint neurosurgical neuroradiological suite may
and pituitary surgery, there have not been reasonable change and improve the overall in-house management and
approaches for vascular and skull base surgery so far [1, 3, workflow of neurovascular emergencies: E.g., currently,
4]. Spatial resolution and lack of practicability of conven- one has to transfer a patient with acute SAH from the
tional angiographic tools did not promote their application in emergency department to the CT scanner, then to the
a surgical environment, despite the clear advantage of e.g. ICU, or to the OR for placement of an external drainage,
having the opportunity to perform direct control of surgical then to the angiography suite for further diagnostics, and
clipping in cerebral aneurysm surgery [5, 6]. Thus, some finally to the operating room, if free at the time. Control
surgical groups perform direct post-operative angiographic angiography has to be scheduled separately in the conven-
control with the patients still anesthetized, in order to take tional setup. The proposed hybrid concept allows for com-
them back to surgery directly in case of incomplete aneu- pletion of all diagnostic and interventional/surgical
rysm clipping or inadvertent clipping of a parent vessel [7]. activities, including CT, DSA, and open, or endovascular
Indocyanine green videoangiography has become a very aneurysm treatment plus immediate control of aneurysm
important tool in microneurosurgical routine [8]. Its particu- occlusion. All can be done in the same room, on the same
lar value concerns the visualisation of small and perforating table, and even jointly (e.g. preparation for arteriography
vessels, whereas vessels, which are not directly visible and placement of external ventricular drainage). Unfortu-
though the microscope may escape microangiographic ana- nately, two patients developed locoregional and revers-
lysis [8]. It should be seen as an excellent complementary ible decubitus after having been placed in the lateral
tool to advanced intraoperative 3D DSA as this is possible position for several hours. It is quite clear that the present
with our proposed concept for integration of FP technology solution with a modern yet standard angiography table is
in a hybrid operating room. Another approach is the use of not sufficient for use in an important number of neurosur-
From Intraoperative Angiography to Advanced Intraoperative Imaging: The Geneva Experience 115
Fig. 5 Intra operative CT images, obtained with Flat Panel system, for resection control in case of large calcified clival chordoma, which was
approached via the sublabial transsphenoidal route. Remnants of the calcified tumour can still be appreciated on axial images (a), as well as
inferiorly on sagittal reconstructions (b)
gical pathologies. The next evolutionary step of such a References

hybrid room should concern the development of a specifi-
cally adapted carbon table, which can be adjusted like any 1. Nimsky C, von Keller B, Schlaffer S, Kuhnt D, Weigel D,
Ganslandt O, Buchfelder M (2009) Updating navigation with
other operating table. intraoperative imaging data. Top Magn Reson Imaging 19:197 204
2. Matula C, Rössler K, Reddy M, Schindler E, Koos WT (1998)
Intraoperative computed tomography guided neuronavigation:
concepts, efficiency, and work flow. Comput Aided Surg
3:174 182
Conclusion 3. Gerlach R, du Mesnil de Rochemont R, Gasser T, Marquardt G,
Reusch J, Imoehl L, Seifert V (2008) Feasibility of Polestar N20,
an ultra low field intraoperative magnetic resonance imaging
The new neuro-interventional suite opens a new avenue for system in resection control of pituitary macroadenomas: lessons
intraoperative imaging by the provision of highly resoluted learned from the first 40 cases. Neurosurgery 63:272 284
CT or angiographic images, which may be co-registered 4. Jankovski A, Francotte F, Vaz G, Fomekong E, Duprez T, Van
with a commercially available navigation system. High tem- Boven M, Docquier MA, Hermoye L, Cosnard G, Raftopoulos C
(2008) Intraoperative magnetic resonance imaging at 3 T using a
poral resolution of vascular diagnostics may assist in devel- dual independent operating room magnetic resonance imaging
oping better virtual and customized treatment plans for suite: development, feasibility, safety, and preliminary experience.
patients suffering from aneurysms and vascular malforma- Neurosurgery 63:412 426
tions. In addition, the workflow in treatment of aneurysmal 5. Alexander TD, Macdonald RL, Weir B, Kowalczuk A (1996)
Intraoperative angiography in cerebral aneurysm surgery: a per
subarachnoid hemorrhage can be improved significantly, spective study of 100 craniotomies. Neurosurgery 39:10 18
as all diagnostic and therapeutic measures can be taken 6. Raabe A, Beck J, Rohde S, Berkefeld J, Seifert V (2006) Three
without having to move the patient to other facilities. In dimensional rotational angiography guidance for aneurysm sur
addition, due to high resolution of bony structures and the gery. J Neurosurg 105:406 411
7. Meyer B, Urbach H, Schaller C, Baslam M, Nordblom J, Schramm
connection with image guidance, there is good potential J (2004) Immediate postoperative angiography after aneurysm
concerning skull base and spinal surgery. The further devel- clipping implications for quality control and further guidance
opment of a fully-functioning carbon-made surgical table is of management. Zentralbl Neurochir 65:49 56
recommended. 8. De Oliveira JG, Beck J, Seifert V, Teixeira MJ, Raabe A (2007)
Assessment of flow in perforating arteries during intracranial
aneurysm surgery using intraoperative near infrared indocyanine
Conflicts of interest statement The project to develop this experi green videoangiography. Neurosurgery 61(3 Suppl):63 73
mental joint interventional suite benefitted from the following: Brain 9. Uhl E, Zausinger S, Morhard D, Heigl T, Scheder B, Rachinger W,
LAB supported the project by the one year cost free use of a Schichor C, Tonn J (2009) Intraoperative computed tomography
navigational system and Philips Medical Systems provided regular with integrated navigation system in a multidisciplinary operating
cost free upgrades of software. suite. Neurosurgery 64(5 Suppl 2):231 240
Intraoperative MRI – Ultra High Field Systems
Intraoperative Magnetic Resonance Imaging
Walter A. Hall and Charles L. Truwit
Abstract Neurosurgeons have become reliant on image- Mapping out eloquent brain function may influence the
guidance to perform safe and successful surgery both time- surgical path to a tumor being resected or biopsied. The
efficiently and cost-effectively. Neuronavigation typically optimal field strength for an ioMRI-guided surgical system
involves either rigid (frame-based) or skull-mounted (frame- and the best configuration for that system are as yet undecided.
less) stereotactic guidance derived from computed tomo-
graphy (CT) or magnetic resonance imaging (MRI) that is Keywords Brain activation Brain neoplasms Intraopera-
obtained days or immediately before the planned surgical tive magnetic resonance imaging Magnetic resonance
procedure. These systems do not accommodate for brain imaging
shift that is unavoidable once the cranium is opened and
cerebrospinal fluid is lost. Intraoperative MRI (ioMRI) sys-
tems ranging in strength from 0.12 to 3 Tesla (T) have been
developed in part because they afford neurosurgeons the Introduction
opportunity to accommodate for brain shift during surgery.
Other distinct advantages of ioMRI include the excellent soft Neurosurgeons have used magnetic resonance imaging
tissue discrimination, the ability to view the surgical site in (MRI) for more than 20 years to visualize the brain because
three dimensions, and the ability to ‘‘see’’ tumor beyond the of the ability to view the brain in three dimensions with
surface visualization of the surgeon’s eye, either with or excellent soft tissue discrimination. Other technological
without a surgical microscope. The enhanced ability to advances that have accompanied MRI over the same time
view the tumor being biopsied or resected allows the surgeon period include frame-based stereotaxy, frameless neurona-
to choose a safe surgical corridor that avoids critical struc- vigation, and most recently, intraoperative MRI (ioMRI)-
tures, maximizes the extent of the tumor resection, and guided neurosurgery. Intraoperative MRI-guidance in the
confirms that an intraoperative hemorrhage has not resulted operating room has the advantage over the other technical
from surgery. Although all ioMRI systems allow for basic advancements in that it allows for the surgeon to visualize
T1- and T2-weighted imaging, only high-field (>1.5 T) the surgical site during the procedure in near-real time. This
MRI systems are capable of MR spectroscopy (MRS), MR capability provides the neurosurgeon with valuable informa-
angiography (MRA), MR venography (MRV), diffusion- tion that will help to guide the surgical procedure to its
weighted imaging (DWI), and brain activation studies. By successful completion primarily by allowing for the com-
identifying vascular structures with MRA and MRV, it pensation of brain shift which occurs after the cranium is
may be possible to prevent their inadvertent injury during opened and cerebrospinal fluid (CSF) is lost [1].
surgery. Biopsying those areas of elevated phosphocholine As ioMRI-guided neurosurgery continues to disseminate
on MRS may improve the diagnostic yield for brain biopsy. to more and more academic health centers and community
hospitals, it has become uncertain which magnetic field
strength is best suited for surgical use. Since the first utiliza-
W.A. Hall (*) tion of this technology in 1994, multiple MRI systems of
Department of Neurosurgery, SUNY Upstate Medical University, varying field strengths have been developed for surgical use
750 East Adams Street, Syracuse, NY 13210, USA
e mail: hallw@upstate.edu
ranging from 0.12 Tesla (T) to 3 T [1, 2]. The first ioMRI
system was a mid-field 0.5 T double coil design (SIGNA SP,
C.L. Truwit
Departments of Radiology, Hennepin County Medical Center, General Electric Medical Systems, Milwaukee, WI) that was
University of Minnesota, Minneapolis, MN, USA installed at Brigham and Women’s Hospital in Boston, MA.
120 W.A. Hall and C.L. Truwit
With this MRI system, the surgeon operated between the considered functional MRI (fMRI) can be obtained at 1.5 T
coils and image updates could be obtained continuously or higher field strength.
during surgery (Fig. 1). Low field MRI systems have ranged Being able to image patients intraoperatively provides
from 0.12 to 0.23 T and high field systems are either 1.5 the neurosurgeon with information that will allow him or
or 3 T in magnet strength. As image quality and MRI her to alter the surgical treatment plan resulting in improved
functionality is proportional to field strength, many diagnos- patient safety and the assurance of successful completion of
tic imaging techniques such as MR angiography (MRA), the procedure. Each field strength ioMRI system has distinct
MR venography (MRV), MR spectroscopy (MRS), diffusion- advantages and disadvantages which are considered by neu-
weighted imaging (DWI), and brain activation studies rosurgeons as they determine whether the various imaging
capabilities provided by each will address their surgical
needs. The surgical applications for each ioMRI system
will be the focus of this contribution.
Low Field ioMRI Systems
The very low field strength 0.12 T ioMRI system (Medtronic

Navigation, Minneapolis, MN) has recently had an increase
in its field strength to 0.15 T (Polestar N20). With this
system the magnets are maintained below the surgical
operating table until imaging is required at which time they
are elevated to each side of the surgical field (Fig. 2). Imag-
ing with this system affords incomplete visualization of the
head due to limitations on field of view, but does allow for
the surgical field of view to be seen. Surgery can be per-
formed entirely within the magnetic field using standard
surgical instrumentation. This ioMRI system allows for con-
Fig. 1 General Electric Signa SP double coil 0.5 T magnetic resonance siderable access to the patient although there are technical
imaging scanner design that was installed at Brigham and Woman’s
Hospital in Boston in 1994. The surgeon is standing directly within the limitations due to the distance between the magnetic poles
scanner and imaging updates can be obtained continuously during the that delimit which patients can be treated. Another advan-
procedure tage of this system is that the surgeon does not need to
Fig. 2 Medtronic PoleStar N20

0.15 T low field intraoperative
MR scanner. The magnetic poles
are elevated into the position for
scanning the head during surgery
(Image courtesy of Medtronic)
Intraoperative Magnetic Resonance Imaging 121
operate between the magnets as with the 0.5 T ioMRI diminish gradient strength and limit the field of view. The
system. decreased field strength found at isocenter of the gap results
Other low field MRI systems that have been used for in reduced SNR, longer imaging times, reduced spatial
diagnostic and ioMRI have a vertical gap or biplanar design resolution, and a limitation of functional and physiological
where the magnet strength is either 0.2 or 0.23 T (Fonar, imaging capabilities. The advantage of this ioMRI system,
Melville, NY; Hitachi Medical, Twinsburg, OH). These greater access to the patient, ultimately was overshadowed
systems have two horizontal 25 40 cm diameter magnetic by the reduced image quality generated at isocenter and the
poles that allow access to the patient from the side. A more broad acceptance of either moving the patient a short
modification to this system resulted in a C-arm design distance in and out of the scanner in a higher field strength
(Siemens Medical Systems, Erlangen, Germany; Marconi scanner or shuttling the scanner in and out of the surgical
Medical, Highland Heights, OH) where a column on one field, as described below.
side of the magnet supported the upper magnetic pole which
allowed for 240 of access to the patient and the ability to
biopsy lesions of the neck and high cervical spine. The
40 cm gap between the magnetic poles again delimits the High Field ioMRI Systems
size of the patients that can be treated with these systems.
Even though these ioMRI systems seem to be more open High field ioMRI systems have the advantage over low and
than short bore cylindrical scanners, during surgery the head mid field systems of advanced functionality including fat
of the patient is as far away from the surgeon as it is with suppression, perfusion studies, MRA, MRV, MRS, MR ther-
high field systems. mometry, and fMRI [1, 3]. One of the initial problems
Initially some centers would use these systems for intrao- identified with operating using mid and high field ioMRI
perative imaging by performing the surgery in another room systems was the inability to use standard surgical instrumen-
and then transporting the patient to the magnet for imaging. tation near the magnet. In order to obtain intraoperative
Others would operate immediately adjacent, but outside the images of the surgical site the patient had to be transported
5 Gauss line, and yet others would work within the magnet into the scanner or the scanner had to be moved to the
itself. Of note, while these ioMRI systems seem to be more patient. Our initial ioMRI system (ACS-NT, Philips Medical
open than short bore cylindrical scanners, it is important to Systems, Best, Netherlands) required that a modified angio-
point out that during surgery on these systems; the head of graphy table serve as the operating table which was rotated
the patient is actually as far away from the surgeon as it is to dock with the MRI gantry so that the patient could be
with high field systems. rapidly transported into and out of the scanner on a floating
All low field systems have reduced temporal resolution, table top. Access to the patient was enhanced with this
spatial resolution per unit time, and reduced signal-to-noise system because of the shore-bore magnet configuration and
ratio (SNR) compared to high field ioMRI systems [1]. 100 cm flared openings on both sides of the scanner. The
These imaging constraints mandate an increase in imaging infrequent need to scan the patient during a brain tumor
time that is sufficient to achieve image quality that will allow resection supports the performance of surgery outside the 5
for accurate surgical decision making. Gauss line using conventional surgical instruments. In con-
trast, brain biopsies were performed within the magnet in
order to enable the visualization of the passage of the biopsy
needle toward the target in near-real time. A comparable
Mid Field ioMRI Systems approach to patient access was adapted at other high field
ioMRI sites (University of Erlangen, Erlangen, Germany
The prototype 0.5 T mid field system was the Signa SP and University of California Los Angeles, Los Angeles,
‘‘double donut’’ ioMRI system that was designed specifically California) where a rotating surgical table could be turned
for interventional use. With this system, two cylindrical into the axis of a 1.5 T Magnetom Sonata Maestro Class
magnets were set up such that an intervening gap was scanner (Siemens Medical Solutions, Erlangen, Germany)
allowed for the surgeon and patient. Each magnet generated [4]. With our updated 1.5 T ioMRI scanner (Intera I/T,
an inhomogeneous magnetic field, which because they were Philips Medical Systems, Best, Netherlands), we performed
inverted relative to each other, created a homogeneous mag- craniotomies for brain tumor resection entirely within the
netic field of 30 cm diameter. This magnet configuration magnetic field using MR-compatible instrumentation and
where the superconducting magnets are separated by a anesthesia equipment.
56 cm gap allows for lateral and vertical access to the Another alternative high field ioMRI system design uti-
imaging isocenter. During procedures, the surgeon stands lizes a mobile 1.5 T magnet with a 92 cm bore diameter
in this gap which requires special radiofrequency coils that (Magnex Scientific, Abingdon, Oxon, UK) that moves on a
ceiling-mounted track to a stationary operating table (Fig. 3). scanning. Initially, this system design did not allow access
This system has the disadvantages that it requires reposition- to the patient during imaging so that minimally invasive
ing local radiofrequency shielding each time an intra- procedures such as brain biopsy were not possible. A more
operative MRI is performed to maintain high SNR and all recent version of the system where the back end of the
ferromagnetic instrumentation must be removed prior to scanner is now open allows for access to the patient so
MR-guided minimally invasive procedures can be per-
formed in near-real time.
Most recently 3 T MRI scanners have been modified to
allow for MR-guided neurosurgical procedures at the back
end of the magnet (Intera, Philips Medical Systems, Best,
The Netherlands) entirely within the magnetic field due to
the greater availability of MR-compatible instruments and
equipment (Fig. 4). Both craniotomies and minimally inva-
sive brain biopsies can be performed with minor limitations
in patient positioning and a significant improvement in
patient flow and throughput due to more rapid acquisition
of intraoperative images [5]. All major vendors have now
modified diagnostic 3 T MR scanners to allow for MR-
guided surgical procedures.
Results
Surgical Indications
The primary indication for 0.15 T ioMRI system is to guide

the resection of brain tumors such as gliomas, pituitary
tumors, and meningiomas. Another group found that their
Fig. 3 IMRIS 1.5 T intraoperative magnetic resonance imaging system 0.2 T low field ioMRI system was best suited for the resec-
with the magnet suspended from ceiling rails. In the foreground is the tions of gliomas and pituitary adenomas. Gliomas and pitui-
MRI compatible surgical table tary adenomas were found to be those lesions that were also
Fig. 4 Philips 3.0 T Intera

intraoperative MRI system where
surgical procedures are
performed at the back end of the
scanner with direct neurosurgical
access to the patient’s head. The
patient can be transported rapidly
into the scanner for frequent
repeat imaging. The column to
the left of the back end of the
scanner serves as the source for
suction and anesthetic gases
most appropriate for 1.5 T MR-guided neurosurgery either Brain Biopsy

with a stationary patient or a stationary magnet.
The initial ioMRI-guided brain biopsies were performed in a
freehand manner much like the first computed tomography-
guided brain biopsies. Clinicians recognized early on that
Brain Shift there was no way to either guide the biopsy needle to the
intended target or to stabilize the needle during the acquisi-
As accurate as neuronavigational systems are, they can suf- tion of the tissue samples. The need for a device to stabilize
fer significant spatial inaccuracy due to the brain shift that the biopsy needle led to the rapid design and development
occurs once the cranium is opened and CSF is drained upon of such instrumentation (MRI Devices, Waukesha, WI;
opening the dura mater. These devices do not allow the Snapper-Stereo-Guide, MagneticVision, Zurich, Switzerland;
neurosurgeon to compensate for brain shift without repeat Image-Guided Neurologics, Melbourne, FL). A disposable
imaging of the operative site and reregistration of those trajectory guide called the Navigus that was combined with a
intraoperative images. The average amount of surface unique targeting technique known as prospective stereotaxy
displacement that has been estimated to occur with the that allowed for the performance of brain biopsy in near-real
resection of a tumor is 1 cm with gravity representing the time at 1.5 T (Fig. 5). Biopsies are generally performed
main force causing brain shift [1]. Peritumoral brain edema under general anesthesia at most ioMRI sites for patient
secondary to the effects of surgery also distorts the local comfort because of the length of the procedure and to pre-
environment with consequent brain shift. This shift can lead vent the inadvertent displacement of the biopsy needle once
to incomplete tumor resection in the face of the neuronavi- the target has been reached.
gation system suggesting complete resection. Alternatively, Prospective stereotaxy is a novel biopsy technique that
there is the potential for surgical encroachment on normal starts at the target and moves away distally toward an align-
brain parenchyma adjacent to the tumor resection cavity, in ment stem that acts similarly to a computer joy stick. With
the face of safe passage suggested by neuronavigation. The this alignment technique three points are aligned: biopsy site
percentage of cases where neuronavigation reregistration (target point), tip of the alignment stem (pivot point), and the
was felt to be necessary in ioMRI-guided cases due to cross section of the alignment stem (rotation point). Once all
brain shift was 11 16% [6]. The occurrence of brain shift three points are collinear, the passage of the biopsy needle
has been in part responsible for causing unsuspected residual through the guide tube which had held the alignment stem
tumor in 2 35% of cases depending on the type of tumor will result in the biopsy needle encountering the target. The
being resected with pituitary adenomas representing the entire alignment process takes 2 5 min and following the
most likely tumor to be incompletely resected. procedure a series of three scans (half-Fourier acquisition
Fig. 5 The skull mounted

Navigus trajectory guide in
position for brain biopsy. The
alignment stem is inserted into the
guide tube and the white plastic
locking nut is visible. The
procedure is being performed
through a gauze wrapped
radiofrequency surface coil,
which combined with its partner
beneath the patient’s head, allows
for parallel imaging, affording
both high signal to noise ratio
and improved imaging efficiency
single-shot turbo spin echo (HASTE), turbo fluid-attenuated In order to enhance the diagnostic yield of brain biopsy,
inversion recovery (FLAIR), and gradient echo (GE)-T2*) we have combined intraoperative guidance with MRS, ini-
are reviewed to exclude the presence of intraoperative tially with single-voxel spectroscopy (SVS) and subsequently
hemorrhage before the conversion of intracellular oxy- with turbo spectroscopic imaging (TSI). Single voxel spec-
hemoglobin to deoxyhemoglobin (Fig. 6). More recently, troscopy targets a region of interest in the brain and com-
susceptibility-weighted imaging has supplanted the imaging pares it to the similar location in the opposite cerebral
protocols due to its increased sensitivity to hemorrhage that hemisphere and TSI is performed on a single axial slice
is still largely in the intracellular oxyhemoglobin phase and where specific brain metabolites (phosphocholine, creatine,
only slightly in the intracellular deoxyhemoglobin phase. N-acetyl aspartate (NAA), and lactate/lipid) are measured
[2]. Regions of elevated phosphocholine on SVS and TSI, in
the presence of diminished NAA, suggest increased cellular
density due to rapid membrane turnover that is consistent
with tumor tissue and N-acetyl aspartate which is a neuronal
marker is increased in normal brain [1].
Craniotomy for Tumor Resection
At our site, patients are intubated and receive general anes-

thesia before they arrive in the ioMRI suite. The head is
secured in the operative position in a carbon fiber head
holder in order to facilitate repeat intraoperative imaging in
the exact orientation that will allow for comparison of scans
during surgery. The craniotomy site is localized over the
tumor using MRI-visible markers before the hair is clipped
and the skin prepped. For nonenhancing brain tumors,
HASTE and turbo FLAIR imaging sequences are acquired
for intraoperative scan comparison; for enhancing tumors,
the administration of intravenous contrast is withheld until
the majority of the tumor has been removed to prevent the
imbibition of contrast into the edematous peritumoral brain
where the blood brain barrier has been disrupted. The inter-
pretation of intraoperative images can be difficult even with
the repeat administration of the intravenous contrast because
of its continued diffusion around the resection cavity that
occurs related to surgical trauma from the use of electrocau-
tery and the ultrasonic aspirator.
Typically, tumor resections require three imaging sets
during the surgery. The first set is acquired prior to the
opening of the cranium (Fig. 7) before brain shift has
occurred. After the tumor has been either partially (Fig. 8) or
completely resected (Fig. 9) in the estimation of the surgeon,
a second, and truly intraoperative set of images is acquired.
Finally, after the craniotomy has been closed, a third set is
obtained, in order to evaluate the surgical site for intraopera-
tive hemorrhage. In addition, at any point during the surgery,
where there could be a question concerning brain physiology
Fig. 6 Following the brain biopsy, a series of three axial scans (half or the extent of resection imaging can be repeated with a
Fourier acquisition single shot turbo spin echo (HASTE) (top), turbo 5 min delay in order to maintain the sterility of the operative
fluid attenuated inversion recovery (FLAIR) (middle), and gradient field and to allow for removal of all non-MR-compatible
echo (GE) T2* (bottom)) are reviewed to exclude the presence of
intraoperative hemorrhage before the conversion of intracellular oxy
instrumentation before transport into the scanner. Imaging
hemoglobin to deoxyhemoglobin has occurred. An area of signal loss in of the operative site in any orientation is possible to enable
the same location on all three scans would indicate the presence of air the surgeon to visualize those areas of the brain that may
Fig. 7 Preoperative axial turbo fluid attenuated inversion recovery Fig. 9 Postoperative axial half Fourier acquisition single shot turbo
(FLAIR) MRI scan demonstrating an area of increased signal in the spin echo (HASTE) scan demonstrating the completion of the tumor
left frontal lobe directly below the coronal suture consistent with a low resection. The entire tumor footprint has been removed radiographi
grade glioma. There is an MRI visible marker attached to the scalp in cally and the resection cavity has been filled with cerebrospinal fluid.
order to localize the craniotomy flap preoperatively and delimit the size There is pneumocephalus still present over the left frontal lobe and the
of the cranial opening and operative field craniotomy has been closed
contain residual tumor that is not easily seen because of the

collapse or ‘‘shift’’ of the brain into or away from the resec-
tion cavity (Fig. 10).
After baseline imaging is obtained, at the 1.5 T site, the
patient is moved outside the magnetic field to the location
where ferromagnetic instruments can be used without risk of
projectile injury. Resection of the tumor is carried out in the
standard neurosurgical fashion until the surgeon feels it
appropriate to assess the completeness of the surgery.
Some groups have combined ioMRI-guided brain tumor
resections with neuronavigation to improve the safety of
the surgery and to increase the overall extent of the tumor
resection [4, 6]. During surgery the imaging data set that is
performed to assess the extent of the tumor resection is
reregistered into the neuronavigational system to further
guide the resection of any residual neoplastic disease.
Because of the additional time associated with combining
neuronavigation with ioMRI-guided surgery, other sites, like
ours, have taken a more simplistic approach for detecting
residual tumor after the initial attempt at removing the
tumor footprint. Prior to obtaining intraoperative images
to examine the surgical site for residual disease, we place
MRI-visible markers in the resection cavity adjacent to
Fig. 8 Oligodendroglioma. Intraoperative axial half Fourier acquisi tissue that is suspected to represent tumor. These markers
tion single shot turbo spin echo (HASTE) scan demonstrating the are made of titanium and are in the shape of rods, struts, or
extent of the tumor resection and the proximity of the residual disease
to the central sulcus and motor cortex. Note the degree of brain shift of
burr hole covers (Fig. 11). Bone wax has also been placed is
the left frontal lobe that is present. This tumor was found on pathologi a resection cavity to monitor for residual disease because of
cal examination to be an oligodendroglioma its excellent visualization characteristics on MRI (Fig. 12).
Fig. 10 Postoperative axial half Fourier acquisition single shot turbo Fig. 12 Axial T1 weighted contrast enhanced MRI demonstrating the
spin echo (HASTE) scan after the left frontal tumor has been resected complete resection of an enhancing left occipital brain metastasis. The
and the craniotomy has been closed. The resection cavity is filled with area of signal loss in the resection cavity represents bone wax that was
cerebrospinal fluid and there is pneumocephalus over both frontal lobes placed adjacent to an area that was felt to be suspicious for residual
demonstrating the degree of ‘‘brain shift’’ that can occur within the disease by the surgeon. Notice the degree of ‘‘brain shift’’ over the left
cranium during surgery despite the tumor being unilateral cerebral hemisphere that is the result of cerebrospinal fluid loss after the
dura mater has been opened
The presence of residual disease near an intraoperative tita-

nium marker will guide the surgeon to that location in order
to extend the tumor resection. Once the tumor footprint has
been removed the three previously described ‘‘hemorrhage’’
scans are obtained before leaving the ioMRI suite. For
tumors adjacent to blood vessels, DWI can be performed at
the time of the hemorrhage scans to exclude the presence
of ischemia or infarction suggestive of a vascular injury.
After imaging is complete, the patient is transported to the
recovery room for extubation. In general, each image set
requires 10 15 min to obtain and they are rarely repeated
within 1 h of the prior imaging. Initially, intraoperative
imaging extended the duration of a tumor resection by one-
third compared to cases performed in the main operating
room, although this time has decreased significantly with
increasing experience using this technology.
Functional-MRI-Guided Tumor Resection
To enhance the overall safety of ioMRI-guided brain

Fig. 11 Intraoperative coronal half Fourier acquisition single shot tumor resection, we have combined fMRI-guidance to the
turbo spin echo (HASTE) scan where a titanium burr hole cover has preoperative surgical planning for tumors located near areas
been placed on the right frontal brain surface to localize an area of of eloquent cortical function. Eloquent cortex represents
increased signal in a single gyrus. The cortical surface appeared normal cortical tissue whereby injury or damage will result in a
under direct visualization. The gyrus was resected and was found to
represent an area of cortical dysplasia that was causing seizures neurological deficit. Brain activation areas that we have
been able to accurately define and avoid during surgery are tumors located adjacent to areas for language activation, the
those for motor function, working memory, and language combination of awake craniotomy using total intravenous
(Fig. 13). Motor cortex activation will result from finger anesthesia and preoperative fMRI language localization has
tapping, tongue tapping on the roof of the mouth, and toe resulted in the safe and successful resection of lesions that
wiggling. Language location is mapped by having patients previously would have been biopsied only. Even though
think of the names of animals starting with the beginning of most fMRI-guided tumor resections have been performed
the alphabet (‘‘silent’’ speech). Short- and long-term recall on high field systems, one group has demonstrated in normal
using list retention is used to localize memory function to the volunteers that brain activation studies are possible on a
medial temporal lobe of the dominant cerebral hemisphere. 0.15 T system [7].
The imaging protocol for fMRI is a single-shot echo planar
imaging (EPI) scan. Those areas of blood oxygen level-
dependent (BOLD) activation are calculated and then
superimposed on high quality anatomic images that can be Diffusion and Perfusion Imaging-Guided
displayed on liquid crystal display panels in the ioMRI Tumor Resection
surgical suite for the surgeon to review immediately prior
to craniotomy. To assure that the task is being properly
Using the diffusion energy (Brownian motion) of water
performed, a superimposed waveform demonstrates those
molecules, the white matter tracts that pass through the
periods where brain activation is occurring. Acquiring and
brain can be mapped with diffusion tensor imaging (DTI).
processing brain activation studies will add approximately
This technique was first used to avoid the optic radiations in
15 min to a surgical procedure provided that they are
a child with a low grade glioma where their inadvertent
obtained immediately before the administration of general
injury would result in a visual field deficit [8]. When com-
anesthesia. Some patients will have their fMRI performed
paring the preoperative and intraoperative DTI tractography
several days before the planned surgical procedure if
in glioma resection surgery, the maximum shifting of the
additional surgical planning is felt to be necessary. In our
white matter ranged from 8 to +15 mm [9]. White matter
experience, fMRI-guidance for tumor resection has been so
tracts shifted outward in 62% of patients and inward in 30%
safe, reliable, and accurate that we have not combined it with
[9]. Intraoperative visualization of the white matter tracts
either awake craniotomy or cortical stimulation. For large
has allowed for the safe resection of gliomas near eloquent
cortex [10].
Very recently perfusion MRI has been used to guide the
resection of residual high grade glioma tissue during ioMRI-
guided surgery. Areas of increased perfusion on ioMRI
corresponded to areas of residual tumor tissue on pathologi-
cal examination confirming the need for additional tissue
resection. These areas of increased perfusion were also
found to correlate with areas of contrast enhancement
when the neuronavigation was updated.
Discussion
Intraoperative MRI-guidance has been felt to improve health

outcomes for patients by reducing the repeat tumor resection
rate, hospital length of stay, and overall hospital costs [11].
High field ioMRI systems clearly offer some advantages
over lower field systems namely the higher functionality
that allows for advanced MRI that includes MRA, MRV,
Fig. 13 Axial turbo fluid attenuated inversion recovery (FLAIR) scan MRS, DWI, and perfusion MRI. This advanced functionality
demonstrating the area of superimposed brain activation immediately has led some investigators to consider whether or not there
posterior to the area of increased signal that was found to be a low grade
oligodendroglioma on pathological examination. The task being per
would be an advantage to ioMRI-guided surgery at 3 T [5].
formed on the blood oxygen level dependent (BOLD) study was tongue We initially adapted a diagnostic 3 T scanner (Philips Medi-
tapping cal Systems) for intraoperative use (Fig. 4) although other
can be advanced rapidly into the magnet at any time imaging

is felt to be necessary. The disadvantages associated with 3 T
ioMRI-guided surgery were the limitations associated with
not having stainless steel instrumentation with its greater
strength over titanium. There is also more pronounced im-
aging artifact associated with the titanium biopsy needle
(Fig. 14). This may also suggest that titanium aneurysm
clips will have significant imaging artifact that makes
image interpretation difficult and it may not be possible to
tell whether there is residual aneurysm neck remaining.
There is also a more pronounced artifact when air is in the
tumor resection cavity which confounds interpreting the
postoperative hemorrhage scans. As at 1.5 T, to minimize
such needle related artifact, it is helpful to align the needle
whenever possible with the main magnetic field Bo. This
works for most intracranial lesions, although is suboptimal
for temporal lobe lesions.
After over a decade of experience, intraoperative MRI-
guided surgery has finally been embraced by the neurosurgi-
cal community as evidenced by five separate systems being
operational in one metropolitan area, two at academic cen-
ters and three in community hospitals. As to whether there is
any advantage to considering magnet strength higher than
Fig. 14 Coronal T1 weighted contrast enhanced MRI scan after the 3 T is a question for debate. It is more likely that 1.5 T and
MRI compatible brain biopsy needle has been passed through the 3 T ioMRI systems will be combined with state-of-the art
Navigus down to the target. The small enhancing focus that was computed tomography, angiography, or positron emission
found to be recurrent tumor is obscured by the artifact associated with
tomography equipment or some combination of such.
the biopsy needle in this orientation. Alignment along Bo when possi
ble will delimit the metallic artifact. In this case, as the lesion was along
the mesial aspect of the left hemisphere, a long axis approach might Conflicts of interest statement We declare that we have no conflict
have likely injured vascular structures as the needle passed through of interest.
multiple sulci
vendors followed suit shortly thereafter, such that there are References
now several 3 T systems modified for surgery. The primary
difference between the 1.5 T and the 3 T ioMRI suites is that 1. Hall WA, Truwit CL (2008) Intraoperative MR guided neuro
at the higher field strength everything that enters the room surgery. J Magn Reson Imaging 27:368 375
2. Hall WA, Truwit CL (2005) 1.5 T:spectroscopy supported brain
must be MRI-compatible. The only instruments that we have biopsy. Neurosurg Clin N Am 16:165 172
not found a non-ferromagnetic replacement for are the scal- 3. Hall WA, Liu H, Martin AJ, Pozza CH, Maxwell RE, Truwit CL
pel blade and the suture needles. The scalpel blade that we (2000) Safety, efficacy and functionality of high field strength
use has a retractable plastic guard that covers the blade when interventional MR imaging for neurosurgery. Neurosurgery
46:632 641
not in use and the MRI-compatible titanium suture needles 4. Nimsky C, Ganslandt O, von Keller B, Romstock J, Fahlbusch R
are both cost prohibitive and less sharp. To the surgeon, (2004) Intraoperative high field strength MR imaging: implemen
when operating in the 3 T environment, there is virtually tation and experience in 200 patients. Radiology 233:67 78
no detectable difference as compared to the 1.5 T field 5. Truwit CL, Hall WA (2006) Intraoperative magnetic resonance
imaging guided neurosurgery at 3 T. Neurosurgery 58(Suppl 2):
strength environment. The advantages that we have seen at ONS338 ONS346
3 T are improved SNR, which provides improved visualiza- 6. Nimsky C, Ganslandt O, Hastreiter P, Fahlbusch R (2001) Intra
tion of vascular structures and enhanced performance of operative compensation for brain shift. Surg Neurol 56:357 365
MRS and fMRI protocols. There has been no demonstrable 7. Azmi H, Biswal B, Salas S, Schulder M (2007) Functional imaging
in a low field, mobile intraoperative magnetic resonance scanner:
increased risk to the patient or staff, while there has been expanded paradigms. Neurosurgery 60:143 149
increased efficiency. This improved efficiency, however, may 8. Tummala RP, Chu RM, Liu H, Truwit CL, Hall WA (2003)
be in part related to the newer configuration whereby the Application of diffusion tensor imaging to magnetic resonance
patient table is extended out the far end of the magnet and guided brain tumor resection. Pediatr Neurosurg 39:39 43
9. Nimsky C, Ganslandt O, Hastreiter P, Wang R, Benner T, Sorensen diffusion tensor imaging based fiber tracking. Neuroimage
AG, Fahlbusch R (2005) Preoperative and intraoperative diffusion 30:1219 1229
tensor imaging based fiber tracking in glioma surgery. Neuro 11. Hall WA, Kowalik K, Liu H, Truwit CL, Kucharczyk J (2003)
surgery 56:130 138 Costs and benefits of intraoperative MR guided brain tumor resec
10. Nimsky C, Ganslandt O, Merhof D, Sorensen AG, Fahlbusch R tion. Acta Neurochir Suppl 85:137 142
(2006) Intraoperative visualization of the pyramidal tract by
3 T ioMRI: The Istanbul Experience
M. Necmettin Pamir
Abstract Intraoperative imaging technologies have impro- Introduction

ved surgical results in glioma and pituitary adenoma sur-
geries. With improvements and refinements 3 T intraoperative The drive for the development of intraoperative imaging
MRI systems offer a potential of further improving these came from the field of neurooncology, due to a need for
results. Hereby we describe the equipment and technique of improving the efficiency of tumor resection. The role of
a cost-effective shared-resource 3-T Ultra-high field intra- surgery in the treatment of glial tumors is one of the contro-
operative Magnetic Resonance Imaging system and report versial issues in neurosugical practice. Glial tumors grow in
our continuing experience on surgical tumor resection. an infiltrative fashion which precludes a true oncological-
A description of the facility design and equipment are resection of these tumors [1]. However; studies have shown
given along with examples from our experience on low- that a more radical resection is associated with longer sur-
grade gliomas and transsphenoidal surgeries. Our facility vival, better quality of life and lower chance of malignant
based on the twin room concept and uses a 3-T Siemens degeneration over time [1 3]. Prospective randomized, mul-
Trio (Siemens, Erlangen, Germany) scanner. The unit consists ticenter trials have shown that neither radiotherapy nor che-
of adjacent but independent MRI and operative suites, which motherapy have a significant effect on patient survival in
are connected by a wide door for ioMRI procedure but are used low-grade gliomas (LGG). Currently the most effective form
as conventional MRI and operative units. Rigid head fixation of treatment for LGG is complete surgical excision [2].
during neurosurgery is achieved with a custom designed 5 pin Studies have also indicated that patients do benefit from
head-rest which also combines a 4þ4 channel head coil. the extent of this resection with better outcome after more
Operation is performed using regular non-MRI compatible complete resections [1 3].
equipment and the patient is transferred to the MRI during the Another drive for the development of intraoperative MRI
procedure using a custom designed floating table. Advanced systems was their potential use in transsphenoidal surgery.
sequences such as diffusion weighted and diffusion tensor Most studies have indicated that residual adenoma tissue is
imaging, MR angiography, MR venography, MR spectro- frequently found after transsphenoidal resections of pituitary
scopy can be performed with no changes in the setup and adenomas. Intraoperative detection of these residuals carried
result in image quality comparable to outpatient scans. the prospect of more complete resections.
The intraoperative 3-T ultra high field MRI unit with the
twin room concept permits both diagnostic outpatient ima-
ging and image guided surgery in the same setting and is a cost
effective solution to afford a highly capable ioMRI system. The Development of ioMRI Technology
Keywords Intraoperative MRI Low grade glioma The adaptation of an MRI scanner into the operating room is
Magnetic resonance imaging Pituitary adenoma Ultra- a fairly new concept and therefore the design and technology
high field MRI of the systems have varied considerably. The first intra-
operative MRI unit that became operational was installed
at Brigham and Women’s hospital in Boston in 1994 [4]. The
M.N. Pamir initial design was nicknamed the ‘‘double-donut’’ and was a
Professor and Chairman, Department of Neurosurgery, Acibadem
0.5-T General Electric magnet (SIGNA SP, General Electric
University, School of Medicine, Inonu Cad, Okur Sok 20, 34742
Kozytagi, Istanbul, Turkey Medical Systems, Milwaukee, WI). In this initial design both
e mail: pamirmn@yahoo.com the patient and the surgeon were inside the magnet where the
132 M.N. Pamir
surgery was performed using all non-ferromagnetic surgical normal parenchyma from residual tumor, which can be
equipment. Similar low field (0.2 0.5 T) designs worldwide problematic even in postoperative imaging, was performed
followed this first application including the Erlangen [5] without technical difficulties [15]. Diffusion weighted ima-
and Toronto Siemens low field systems (Magnetom OPEN; ging was used to monitor surgically induced ischemia [15].
Siemens Medical Systems, Erlangen, Germany) and the Previous studies using low field imaging had already shown
Hitachi [6] system (Fonar, Melville, NY; Hitachi Medical, that complications such as intraoperative bleeding could
Twinsburg, OH). The short to midterm follow-up results of be demonstrated using ioMRI [18]. With the use of high
surgeries performed using low field systems have recently and ultra-high field ioMRI intraoperative Diffusion Tensor
begun to appear in the neurosurgical literature [7, 8]. Ana- Imaging (DTI) was also performed to demonstrate the corti-
lyzing the results for glioma surgery with the initial double cospinal tract [19, 20]. All of these technologies are still
donut design from Brigham and Women’s hospital, Claus being validated, further investigated and undergoing refine-
et al. [7] have shown that the odds ratio for recurrence is 1.4, ment and their impact on long-term patient outcome and
and the odds ratio for death is 4.9 when subtotal vs. total quality of life still needs to be investigated.
resection using ioMRI was compared. While some designs aimed at adopting more sophistica-
The adoption of the MRI scanner into the operating room ted magnet designs into the operating room, alternative
was a significant development in the field of neurooncology, strategies have emerged which aimed at developing low-
which also sparked the start of a rapid evolution of this cost, low-bulk systems, exemplified by the 0.12-T system
technology. The use of high-field MRI equipment in the which had permanent magnets of 40 cm diameter placed
operating room was the second important development in 25 cm apart from each other [21]. Use of a 0.15-T version
this field. The image quality and signal-to-noise ratio of the of ultra-low field designs for LGG surgery was reported by
MRI improves with increasing magnet strength and magne- Johann Wolfgang von Goethe University in Germany (Pole-
tic gradients. Low-field systems suffer from low anatomical star N20 system, Medtronic Navigation, Louisville, CO) [8].
resolution, slow speed and the lack of most sophisticated Along with this evolution in magnet technology there was
MRI sequences, which are indispensable in today’s MRI also a significant refinement in operating room designs. The
technology. Application of the 1.5 T technology improved initial Signa SP system was designed to be dedicated opera-
anatomical resolution in imaging; enabled acquisition of tive equipment [4]. The design was based on the concept of
good quality T2 weighted images, which are the standard surgery inside the magnetic field and therefore required the
MRI sequence for low grade glioma. Several high-field use of non-ferromagnetic equipment, which significantly
systems were installed worldwide including the Philips sys- increased the installation and running costs of such a system.
tem at University of Minnesota [9], Siemens systems at These costs increase further with increasing magnet
University of Erlangen [10] and University of California strength. Financing such a big system is a major burden for
and the IMRIS system at Calgary-Canada [11]. Clinical most academic or private institutions. An alternative was
reports have documented that these systems were capable offered by Siemens [5] and Hitachi [6] low-field systems
of increasing the extent of resection [12, 13]. by designing shared-resource facilities where the MRI gan-
Further evolution in intraoperative MRI technology try and the operating room were separated and the MRI was
brought ultra-high field (3 T) scanners into the operating used both for operative and outpatient purposes. Similar
room. This upgrade from 1.5 to 3 T scanners brings further designs were developed for high-field systems [22]. Siemens
refinements in imaging quality and speed. Few ultra-high ioMRI design reverted to single dedicated operating room in
field systems have been reported in the literature so far and the Siemens 1.5-T Brain Suite design [10, 23]. The Siemens
the technology was pioneered by the 3 T-Philips system at system installed at our institution in Istanbul-Turkey used
University of Minnesota-USA [9, 14], the 3 T-Siemens at the twin-room philosophy to reduce cost by allocating the
Acibadem University-Turkey [15, 16] and finally the second equipment to diagnostic outpatient imaging at times when
Philips system at Cliniques Universitaires St-Luc, Université MRI is not needed intraoperatively [15, 16, 24]. A similar
Catholique de Louvain-Belgium [17]. The first clinical design was also adopted by the Philips system installed in
series analyzing the results of low grade glioma surgery Belgium [17].
was reported by our institution and showed an increased
gross total resection rate comparable to the results obtained
by low- and high- field systems [15]. Probably the most
The Acibadem University ioMRI Facility
significant gain from installation of the ultra-high field sys-
tems was the routine application of very high resolution T2 Design
weighted images as the standard imaging sequence. Other
exciting developments also followed: Application of proton The Acibadem University-ioMRI suite brings together a
MR spectroscopy (MRS) and diffusion weighted imaging Siemens Trio 3 T MRI scanner and a twin-room facility
(DWI) to differentiate peritumoral T2w MRI changes in design. This shared resource 3 T-ioMRI setup is not the
3 T ioMRI: The Istanbul Experience 133
adoption of an existing facility but a completely novel imaging to exclude vascular complications, single voxel
design. To complete the system several additional pieces of Proton MRS to differentiate between residual tumor and
equipment were specially designed which include a 5 pin edema, MRA and MRV to demonstrate vascular structures
head-fixation device/4+4 channel head coil combination and were performed as needed. An update of neuronavigation
a floating-patient transfer table. The unit became operational was performed when necessary which took an additional
in June 2004 and after initial optimization of the setup the 3.5 min. Results were interpreted together by the radiologist
first ioMRI procedure was performed in November 2004. and the neurosurgeon.
Within the first 3 years of use a wide array of neurosurgical Early postoperative MRI has been proven effective in
procedures were successfully performed in the 3 T-ioMRI accurately evaluating surgical results in glioma surgery
environment. During this first 3 years a total of 19,217 [25]. The image quality provided by the current high-field
diagnostic scans were performed using the same MRI equip- and ultra-high field ioMRI systems is not any inferior to
ment. Most commonly performed operations at our institu- postoperative MRI studies [14 17, 24, 26]. Therefore, an
tion were tumor resections and transsphenoidal explorations. optimal evaluation of the surgical outcome can be obtained
Other procedures such as brain biopsies, deep brain stimula- during the procedure and further improvements can be
tions, aneurysm and epilepsy surgery were also performed. carried out before the procedure ends without increasing
The advantages of the shared-resource Siemens Trio 3 T morbidity.
system at Acibadem University are several fold (1) This is Although morphologic ioMR images provide valuable
an ultra high-field system with very high image quality, data on the extent of resection, in certain circumstances
advanced imaging capabilities and high speed. (2) The sys- conventional sequences fall short of differentiating treat-
tem is based on the twin room concept which, by allocating ment effects from residual tumor [27]. On preoperative
the system both to intraoperative and outpatient diagnostic MR imaging a significant proportion of the glial tumors
use, therefore offering economically very appealing solution present with a T2w hyperintense rim around the tumor
for institutions which find it difficult to finance a dedicated mass that is suggestive of peritumoral edema or tumor infil-
operating room MRI equipment. MRI compatible surgical tration. In our experience this area expanded slightly after
instruments are not required and neurosurgery is performed resection and this unexpected peritumoral intensity change
using regular ferromagnetic equipment, which is another prompted further studies. In such instances we used Proton
factor that limits the running costs. Magnetic Resonance Spectroscopy (MRS) in combina-
tion with diffusion weighted imaging (DWI) to identify the
nature of this tissue around the resection cavity [15].
When our facility was designed there was no previous
Image Quality: Imaging Capabilities experience with the application of a 3 T magnet to intrao-
perative imaging and therefore we carried concerns that
The 3 T MRI (Trio; Siemens, Erlangen, Germany) is a image distortion and artifact due to increased susceptibility
commercially available, fully functional diagnostic scanner could limit image quality. After optimization of the setup the
that is equipped with strong imaging gradients (40 mT/m, only limitation was in the application of diffusion tensor
200 mT/m/ms) and state-of the art pulse sequences. The EPI- imaging [15]. Analysis of this problem indicated to field
capable gradient set and Syngo 2004A software platform inhomogenities created by ferromagnetic parts in the head
permits fMRI studies, perfusion/diffusion weighted imaging coil. These problems have been solved by modifications in
and proton spectroscopy. The images were reconstructed the head coil and DTI images are now flawlessly acquired.
and displayed immediately after acquisition in the monito-
ring room.
Our routine 3 T- ioMRI protocol in both low grade glioma
and pituitary adenoma patients is based on triplanar very Twin Room/Shared Resource Design
high resolution T2W images. Such very high resolution
images were not possible with lower magnet strengths. Advanced technology certainly has its price: Ultra-high field
These intraoperative images are interpreted by direct com- systems are more expensive than high and low field equip-
parison corresponding baseline images obtained at the start ment. Financing such a big system is a major burden for
of the operation. Baseline images increase the accuracy and most academic or private institutions. But the design of our
greatly simplify the interpretation of intraoperative images. facility greatly reduced this burden by allocating the equip-
Intravenous contrast was not a part of routine baseline ment to diagnostic outpatient imaging at times when MRI is
images, however it was used in certain high grade glioma not needed intraoperatively. When the scanner is not used
or pituitary macroadenoma cases on the on the first-look for intraoperative imaging the radiology team performed an
ioMRI session upon special indication. Diffusion weighted average of 25 outpatient MRI sessions per day and more than
134 M.N. Pamir
19,000 outpatient MRI procedures were performed in the Our pre-ioMRI experience indicated on the important
first 3 years of the facilities use [15]. role of extensive surgical resection in the treatment of low
In the Acibadem University ioMRI suite the MRI gantry/ grade gliomas [31]. The first clinical analysis of low grade
control room and the operating theatre are totally indepen- glioma patients using ultra-high field ioMRI reported by our
dent and for ioMRI sessions the patient is transferred from institution also concluded that the ioMRI detected residual
the operating table into the gantry on a custom designed tumor in approximately half of the patients and this finding
floating table. This 3 m long transfer takes an average of prompted further tumor resection. In 40% of the cases these
1.5 min and the routine imaging paradigm another 6 8 min residuals could not be detected by visual inspection or by
totaling to 10 min for each ioMRI imaging session. This new intraoperative ultrasonography. The use of ioMRI led to an
design neither causes a significant difference in the daily increase in the number of gross total resections by one third,
number of outpatient diagnostic studies nor blocks an bringing it up to almost 75% (Fig. 1).
operating theater only for ioMRI procedures. Former studies Use of the ioMRI was also important for safety of the
have reported patient transfer as a drawback, but the newly glioma surgery. Cortical mapping studies in patients with
designed operation-transfer table setup has greatly simpli- hemispheric gliomas have demonstrated that these infiltra-
fied and facilitated the process. During the procedure the tive lesions may contain functionally normal tissue within
sliding surface of the patient table is conveniently moved to the tumor substance [26]. Resection of gliomas in functional
the docked transfer table and the patient enters the gantry on brain regions caries a risk of injury to neighboring cortical
this floating transfer table. As the head rest is attached to the areas and/or subcortical pathways. Precise and accurate
sliding surface, the patient position does not change. The information on cortical function (from fMRI) and subcor-
transfer process is simple and therefore less prone to techni- tical connectivity (from DTI) is of great importance for
cal failure. In shared resource designs the gantry room is protection of valuable brain tissue and preserving functional
used both for operative and outpatient purposes and risk of integrity. The information helps in both defining resection
infection is cited as a potential problem [5, 6]. In the Aciba- margins and also surgical planning to choose safe entry
dem University ioMRI system the operating room is pres- routes. Our report on low grade gliomas has indicated that
sure regulated, the gantry room is sterile cleaned before each close to two thirds of the cases were located in eloquent
procedure and the wound is sterile draped during each trans- areas and preoperative imaging indicated presence of func-
fer. We have not encountered a single case of infection tional tissue inside the tumor tissue in one tenth of cases
during the first 6 years of use. No anesthesia related pro- [15]. Resection of the involved area was not attempted such
blems or complications were encountered either. cases. Also, operative complications such as hematomas,
In the shared resource design there is no requirement ischemia and infarction were effectively monitored using
for MRI compatible operative instruments. Regular ferro- both anatomical imaging data as well as DWI.
magnetic neurosurgical instruments, operative microscopes,
navigation setup, diagnostic equipment such as ultrasound,
electrophysiological monitoring and anesthesia equipment
are routinely used during the operation. Compatibility with Application in Transsphenoidal Surgery
regular neurosurgical instruments helps in limiting the
running costs. Only exceptions are the MRI compatible The use of the intraoperative MRI during transsphenoidal
head-rest and MRI compatible anesthesia equipment that is surgery is another major focus of interest in our institution.
used inside the gantry. Our previous experience showed that early postoperative
imaging was very important in determining the surgical
outcome after transsphenoidal surgery for pituitary adeno-
mas [32]. We have recently analyzed our initial results with
ioMRI during transsphenoidal surgery. During the initial 4
Application in Low Grade Glioma Surgery years ater the setup of our facility 42 patients underwent
transsphenoidal resections for nonsecretory adenomas.
Several studies have consistently indicated that the extra 21.4% of these cases were previously operated recurrent
information provided by the low or high field ioMRI aids cases. Our surgical aim was total surgical excision. In
in optimizing high and low grade glioma resections [3, 6 8, cases with frank cavernous sinus invasion the the surgical
12, 13, 15, 22 24, 28 30]. The studies analyzed both aim was excision of the extrcavernous portion followed by
low and high grade gliomas and the incidence of residual radiosurgery for the intracavernous portion. The ioMRI pro-
tumor in these reports ranged from 36 to 94% and further cedure took less than 10 min including the transfer and
resection resulted in gross total tumor removal in an addi- very high resolution T2w images are obtained without the
tional 29 41% of these patients. use of contrast (Fig. 2). Intraoperative MRI findings were
Fig. 1 A left frontal low grade

glioma (a) was operated using
3 T Ultra High Field
intraoperative MRI. The first
intraoperative scan performed
after initial resection and
ultrasonography control revealed
tumor remnant at the posterior
margin of the resection cavity
(b), which was further resected to
yield a gross total resection at the
end of the operation (c). A follow
up MRI 3 months after surgery
confirms gross total resection (d)
Fig. 2 Pituitary macroadenoma

w|th suprasellar extension (as can
be seen on coronal (a) and sagittal
(b) contrast enhanced images)
was operated using 3 T ultra high
field intraoperative MRI.
Intraoperatively acquired coronal
(c) and sagittal (d) high resolution
T2 weighted images revealed
total resection. Identical coronal
(e) and sagittal (f) T2 weighted
images performed at the third
month follow up study confirmed
the findings from intraoperative
images
confirmed at the first follow-up examination 3 months after attempted. At the first-look ioMRI total resection was con-
surgery. In cases with frank cavernous sinus invasion resec- firmed in 69%. In 6.9% no further resection was attempted.
tion of the extracavernous portion was attempted. A first- A re-exploration was prompted in 24.1%, based on residual,
look MRI led to further resection in 3 of 13 cases and further potentially resectable residual adenoma. In 57% of these,
tumor resection was achieved in all re-explorations. In cases a total resection could be achieved which increased the
with no cavernous sinus invasion a total resection was overall total resection rate to 82.8%. In one patient the
136 M.N. Pamir
intraoperative MRI detected a hematoma in the resection 8. Senft C, Seifert V, Hermann E, Franz K, Gasser T (2008) Useful
bed, which was managed with immediate evacuation and ness of intraoperative ultra low field magnetic resonance imaging
in glioma surgery. Neurosurgery 63:257 266, discussion 266 267
hemostasis without resultant morbidity. 9. Hall WA, Liu H, Martin AJ, Pozza CH, Maxwell RE, Truwit CL
(2000) Safety, efficacy, and functionality of high field strength
interventional magnetic resonance imaging for neurosurgery. Neu
rosurgery 46:632 641, discussion 641 642
10. Nimsky C, Ganslandt O, Fahlbusch R (2005) Comparing 0.2 tesla
Conclusions with 1.5 tesla intraoperative magnetic resonance imaging analysis
of setup, workflow, and efficiency. Acad Radiol 12:1065 1079
11. Kaibara T, Saunders JK, Sutherland GR (1999) Utility of a move
• This is a novel intraoperative MRI system and com- able 1.5 Tesla intraoperative MR imaging system. Can J Neurol Sci
bines the shared resource design with an Ultra-high field 26:313 316
scanner. 12. Hatiboglu MA, Weinberg JS, Suki D, Rao G, Prabhu SS, Shah K,
• The shared resource design, and use of normal surgi- Jackson E, Sawaya R (2009) Impact of intraoperative high field
magnetic resonance imaging guidance on glioma surgery: a
cal equipment and instruments makes the design cost prospective volumetric analysis. Neurosurgery 64:1073 1081, dis
effective. cussion 1081
• 3 T-ioMRI was effectively used in neurooncology with 13. Nimsky C, Fujita A, Ganslandt O, Von Keller B, Fahlbusch R
the most dramatic results in low grade glioma resections (2004) Volumetric assessment of glioma removal by intraoperative
high field magnetic resonance imaging. Neurosurgery 55:358 370,
and transsphenoidal surgery discussion 370 371
• Advanced imagines sequences such as Proton MR Spec- 14. Hall WA, Galicich W, Bergman T, Truwit CL (2006) 3 Tesla
troscopy (MRS), diffusion weighted imaging (DWI) dif- intraoperative MR imaging for neurosurgery. J Neurooncol
fusion tensor imaging (DTI) were effectively used 77:297 303
15. Pamir MN, Ozduman K, Dincer A, Yildiz E, Peker S, Ozek MM
intraoperatively. Intraoperative Proton MR Spectroscopy (2009) First intraoperative, shared resource, ultrahigh field 3 Tesla
(MRS) was useful in differentiating peritumoral paren- magnetic resonance imaging system and its application in low
chymal changes from residual tumor tissue. grade glioma resection. J Neurosurg
16. Pamir MN, Peker S, Ozek MM, Dincer A (2006) Intraoperative
Conflict of Interest Statement We declare that we have no conflict MR imaging: preliminary results with 3 tesla MR system. Acta
of interest. Neurochir Suppl 98:97 100
17. Jankovski A, Francotte F, Vaz G, Fomekong E, Duprez T, Van
Boven M, Docquier MA, Hermoye L, Cosnard G, Raftopoulos C
dual independent operating room magnetic resonance imaging
References suite: development, feasibility, safety, and preliminary experience.
Neurosurgery 63:412 424, discussion 424 426
18. McClain CD, Soriano SG, Goumnerova LC, Black PM, Rockoff
1. Piepmeier JM (2009) Current concepts in the evaluation and man MA (2007) Detection of unanticipated intracranial hemorrhage
agement of WHO grade II gliomas. J Neurooncol 92:253 259 during intraoperative magnetic resonance image guided neuro
2. Keles GE, Lamborn KR, Berger MS (2001) Low grade hemi surgery. Report of two cases. J Neurosurg 106:398 400
spheric gliomas in adults: a critical review of extent of resection 19. Nimsky C, Ganslandt O, Hastreiter P, Wang R, Benner T, Sorensen
as a factor influencing outcome. J Neurosurg 95:735 745 AG, Fahlbusch R (2007) Preoperative and intraoperative diffusion
3. Sanai N, Berger MS (2008) Glioma extent of resection and its tensor imaging based fiber tracking in glioma surgery. Neurosur
impact on patient outcome. Neurosurgery 62:753 764, Discussion gery 61:178 185, discussion 186
264 266 20. Nimsky C, Ganslandt O, Merhof D, Sorensen AG, Fahlbusch R
4. Schwartz RB, Hsu L, Wong TZ, Kacher DF, Zamani AA, Black (2006) Intraoperative visualization of the pyramidal tract by
PM, Alexander ER, Stieg PE, Moriarty TM, Martin CA, Kikinis R, diffusion tensor imaging based fiber tracking. Neuroimage 30:
Jolesz FA (1999) Intraoperative MR imaging guidance for intracra 1219 1229
nial neurosurgery: experience with the first 200 cases. Radiology 21. Hadani M, Spiegelman R, Feldman Z, Berkenstadt H, Ram Z
211:477 488 (2001) Novel, compact, intraoperative magnetic resonance
5. Steinmeier R, Fahlbusch R, Ganslandt O, Nimsky C, Buchfelder M, imaging guided system for conventional neurosurgical operating
Kaus M, Heigl T, Lenz G, Kuth R, Huk W (1998) Intraoperative rooms. Neurosurgery 48:799 807, discussion 807 809
magnetic resonance imaging with the magnetom open scanner: 22. Ramina R, Coelho Neto M, Giacomelli A, Barros EJ, Vosgerau R,
concepts, neurosurgical indications, and procedures: a preliminary Nascimento A, Coelho G (2009) Optimizing costs of intraoperative
report. Neurosurgery 43:739 747, discussion 747 748 magnetic resonance imaging. A series of 29 glioma cases. Acta
6. Bohinski RJ, Kokkino AK, Warnick RE, Gaskill Shipley MF, Neurochir
Kormos DW, Lukin RR, Tew JMJ (2001) Glioma resection in a 23. Maesawa S, Fujii M, Nakahara N, Watanabe T, Saito K, Kajita Y,
shared resource magnetic resonance operating room after optimal Nagatani T, Wakabayashi T, Yoshida J (2009) Clinical indications
image guided frameless stereotactic resection. Neurosurgery for high field 1.5 T intraoperative magnetic resonance imaging and
48:731 742, discussion 742 744 neuro navigation for neurosurgical procedures. Review of initial 100
7. Claus EB, Horlacher A, Hsu L, Schwartz RB, Dello Iacono D, cases. Neurol Med Chir (Tokyo) 49:340 349, discussion 349 350
Talos F, Jolesz FA, Black PM (2005) Survival rates in patients with 24. Pamir MN, Ozduman K (2009) 3 T ultrahigh field intraoperative
low grade glioma after intraoperative magnetic resonance image MRI for low grade glioma resection. Expert Rev Anticancer Ther
guidance. Cancer 103:1227 1233 9:1537 1539
25. Ekinci G, Akpinar IN, Baltacioglu F, Erzen C, Kilic T, Elmaci I, 29. Nimsky C, Ganslandt O, Buchfelder M, Fahlbusch R (2006) Intrao
Pamir N (2003) Early postoperative magnetic resonance imaging perative visualization for resection of gliomas: the role of function
in glial tumors: prediction of tumor regrowth and recurrence. Eur J al neuronavigation and intraoperative 1.5 T MRI. Neurol Res
Radiol 45:99 107 28:482 487
26. Truwit CL, Hall WA (2006) Intraoperative magnetic resonance 30. Ntoukas V, Krishnan R, Seifert V (2008) The new generation
imaging guided neurosurgery at 3 T. Neurosurgery 58:ONS 338 polestar n20 for conventional neurosurgical operating rooms: a
ONS45, discussion ONS 345 ONS 346 preliminary report. Neurosurgery 62:82 89, discussion 89 90
27. McBride DQ, Miller BL, Nikas DL, Buchthal S, Chang L, Chiang 31. Kilic T, Ozduman K, Elmaci I, Sav A, Necmettin Pamir M (2002)
F, Booth RA (1995) Analysis of brain tumors using 1H magnetic Effect of surgery on tumor progression and malignant degeneration
resonance spectroscopy. Surg Neurol 44:137 144 in hemispheric diffuse low grade astrocytomas. J Clin Neurosci
28. Duprez TP, Jankovski A, Grandin C, Hermoye L, Cosnard G, 9:549 552
Raftopoulos C (2008) Intraoperative 3T MR imaging for spinal 32. Kilic T, Ekinci G, Seker A, Elmaci I, Erzen C, Pamir MN (2001)
cord tumor resection: feasibility, timing, and image quality Determining optimal MRI follow up after transsphenoidal surgery
using a ‘‘twin’’ MR operating room suite. Am J Neuroradiol 29: for pituitary adenoma: scan at 24 hours postsurgery provides reli
1991 1994 able information. Acta Neurochir 143:1103 1126
Intra-operative 3.0 T Magnetic Resonance Imaging
Using a Dual-Independent Room: Long-Term Evaluation
of Time-Cost, Problems, and Learning-Curve Effect
X. Pablos Martin, G. Vaz, E. Fomekong, G. Cosnard, and C. Raftopoulos
Abstract We present a short and comprehensive report of Introduction

our 39-month experience using a 3.0 T intra-operative mag-
netic resonance imaging (ioMRI) neurosurgical-MR twin The first system to use intra-operative Magnetic Resonance
room, including a description of the problems encountered Imaging (ioMRI) for neurosurgical procedures was built in
and the associated time-delays. Forty-seven problems were the Brigham and Women’s Hospital in Boston in 1994 [1].
experienced during the 189 ioMRI procedures (two ioMRI Since then, a wide range of ioMRI models has been deve-
were performed in five of the 184 surgical procedures) loped, each with its own advantages and limitations (see [2]
performed in the 39-month period, including a blocked for a review). Regardless of the type of ioMRI system used,
transfer table, failure of anesthetic monitoring material, operating room time has been shown to be increased [3 12]
and specific MRI-related problems, such as head and coil (but see [13]). The additional time needed to perform an
positioning difficulties, artefacts, coil malfunctions and ioMRI procedure is closely related to the configuration of
other technical difficulties. None of these problems pre- the specific ioMRI system, and to the occurrence of technical
vented the ioMRI procedure from taking place or affected difficulties. Based on our 39-month experience with the
image interpretation, but they sometimes caused a signifi- ioMRI dual-independent room suite built in our academic
cant delay. Fifteen (32%) of these problems occurred during hospital [14], we present the first detailed report of the time-
the initial learning curve period. The mean duration of the cost associated with ioMRI and the technical difficulties
ioMRI procedure was 75 min, which decreased slightly with encountered. This report provides useful information to any-
experience, although an average waiting-for-access time of one interested in advanced ioMRI development.
24 min could not be avoided. These results illustrate that
although performing ioMRI at 3.0 T with the dual room is a
challenging procedure, it remains safe and feasible and
associated with only minor dysfunctions while offering
optimal image quality and standard surgical conditions.
Keywords Intra-operative magnetic resonance imaging Between February 2006, when the first ultra high-field
Neuroimaging Neurosurgery ioMRI was performed at our institution, and May 2009,
180 patients underwent 184 surgical procedures using the
3.0 T. ioMRI complex (Table 1; five patients had two ioMRI
so that 189 ioMRI were performed). The mean age of these
patients at the time of surgery was 44 years, ranging from
1.6 to 81 years. The cohort consisted of 88 female patients
(mean age 43 years, range 1.6 to 80 years) and 92 male
patients (mean age 45 years, range 2 to 81 years). Among
X. Pablos Martin, G. Vaz, E. Fomekong, and C. Raftopoulos (*) the 184 surgical procedures, 153 were a first operation
Department of Neurosurgery, Cliniques Universitaires St Luc, and 31 were repeated surgery for tumor progression. Histo-
Université Catholique de Louvain, Avenue Hippocrate, 10, Brussels pathological examination and clinical evaluation revealed
1200, Belgium
that the lesions were glioma (n¼64, WHO grade I: n¼11;
G. Cosnard
Department of Radiology, Cliniques Universitaires St Luc,Université grade II: 9; grade III: 15; and grade IV: 29), pituitary
Catholique de Louvain, Brussels Belgium adenoma (n¼48), meningioma (n¼14), metastasis (n¼14),
140 X.P. Martin et al.
Table 1 Characteristics of patients who underwent ioMRI between Feb 2006 and May 2009
Tumors and other pathology Patients, n (%) Sex F/M Age (min max) Primary/recurrent Surgery
Primary
Glioma 64 (46) 29/35 47 (1.6 81) 59/9 68a
b
WHO grade I 11 (6) 4/7 27 (2 68) 11/0 11
WHO grade IIc 9 (5) 7/2 40 (25 67) 9/0 9
WHO grade III 15 (8) 7/8 38 (26 49) 14/4 18
WHO grade IV 29 (16) 11/18 62 (35 81) 25/5 30
Pituitary adenomad 48 (27) 19/29 48 (15 77) 39/9 48
Meningioma 14 (8) 11/3 59 (35 80) 12/2 14
Schwannoma 3 (2) 1/2 68 (60 75) 3/0 3
Metastasis 14 (8) 9/5 54 (25 78) 10/4 14
Cavernoma 3 (2) 3/0 38 (30 44) 1/2 3
Medically refractory epilepsy
Without lesion 13 (7) 7/6 15 (3 29) 13/0 13
With lesione 7 (4) 3/4 15 (4 37) 5/2 7
Otherf 14 (8) 6/8 31 (2 63) 11/3 14
Total 180 88/92 44 (2 81) 153/31 184
a
Including 4 surgical procedures for tumor progression (3, 6, 15 and 17 months)
b
Patient #145: LG glioma or DNET, Patient#147 glioma vs. dysplasia
c
Including three gangliogliomas
d
For six patients, normal tissue was found on histological examination
e
Lesions were: Tubers, dysembryoplastic neuroepithelial tumor (n 2), neuronal dysplasia, pilomyxoid astrocytoma, glioneuronal malformation,
and astroblastoma GII
f
Colloid cyst, choroid plexus papilloma, hemangioblastoma (n 3), liponeurocytoma, medulloblastoma (n 4), chordoma, hemangioma, crano
phayngioma, and epidermoid cyst
medulloblastoma (n¼4), schwannoma (n¼3), cavernoma Results

(n¼3), hemangioblastoma (n¼3), colloid cyst, choroid plex-
us papilloma, liponeurocytoma, chordoma, hemangioma, There were no adverse events related to ferromagnetic instru-
cranopharyngioma, and epidermoid cyst (n¼1 each). In 20 mentation in the MRI room during the study period, and
other patients, a deconnection and/or resection procedure no patient or staff member experienced an injury or was
was performed for treatment of medically refractory epilepsy placed at risk because of the projectile effect. There were
(MRE); of these 20 patients, 13 had cryptogenic epilepsy and no accidents related to patient transfer between the operating
for seven patients, histological analysis revealed a dysem- room (OR) and the MRI room.
bryoplastic neuroepithelial tumor (DNET, n¼2), Bourne- Table 2 summarizes the problems that were encountered
ville’s tubers, neuronal dysplasia, pilomyxoid astrocytoma, while using the ioMRI. Each problem was immediately
astroblastoma (WHO grade II), and glioneuronal malforma- resolved in collaboration with technical and radiological
tion (n¼1 each). staff. None of these problems led to cancellation of the
The design of the ioMRI suite, including imaging cap- procedure or inability to interpret the MRI.
abilities, architectural setup and surgical equipment, has The most frequent problem was a blocked transfer table
been described in detail elsewhere [14]. (n¼13), which occurred mostly during the first 7 months of
For every procedure, data were collected prospectively the study period, with a frequency of one in six ioMRI
concerning the time needed to complete the different ioMRI procedures. Technical inspection of the table revealed an
steps, and problems that occurred during the process, and internal software fault that was easily resolved. This pro-
any consequences of such problems. The data were then blem did not recur over the subsequent 5 months and then
integrated into a spread sheet program for further analysis. reappeared, occurring about once in every 21 ioMRI proce-
To evaluate any possible learning-curve effect, the 189 dures (1/4.33 months). The problem was generally related to
ioMRI procedures were grouped; groups of equal numbers an internal software issue that was solved by restarting the
of procedures rather than of equal periods of time were system. The actual delay caused by the blocked transfer table
formed since the learning experience may be related more was reported in only two cases, as 10 and 17 min. In the other
to the number of procedures performed than to the period cases, the difference between the actual transfer duration and
of time. the average normal transfer duration was estimated as 15 min.
Intra-operative 3.0 T Magnetic Resonance Imaging Using a Dual-Independent Room 141
Table 2 Problems associated with the ioMRI system and their evolution over time
Wait for access to Other problems
ioMRI
ioMRIa Period (month) Cases Av. time Blocked Anesthetist MRI related problems
no. (%) (min) transfer related Unplugged coil Missing or non
table problemsb and coil position functional coil Head position Otherc
21 3 (Feb May 06) 18 (86) 17 3 0 2 0 2 1
21 4 (May Sept 06) 20 (95) 20 4 2 0 0 0 1
21 5 (Sept 06 Jan 07) 21 (100) 25 0 0 1 0 0 1
21 5 (Jan Jun 07) 20 (95) 21 1 1 0 1 1 0
21 4 (Jun Oct 07) 21 (100) 30 1 1 0 0 1 2
21 5 (Oct 07 Mar 08) 20 (95) 27 1 1 0 1 0 1
21 4 (Mar Jul 08) 13 (62) 19 1 1 1 0 2 1
21 4 (Jul Nov 08) 16 (76) 27 1 1 1 0 1 0
21 6 (Nov 08 May 09) 18 (86) 28 1 0 0 5 0 1
189 40 (Feb 06 May 09) 167 (88) 24 13 7 5 7 7 8
a
The 189 ioMRI (two in five patients) were clustered in nine consecutive groups
b
Anesthetist related problems included: respirator (n 2), SPO2, monitoring material problem (n 3), unknown
c
Other MRI related problems included: Metal artifact (n 2); air artefact, neuronavigation registration problem; MRI software bug; air equilibra
tion problem; non MRI adapted tube; second contrast injection and new MRI
MRI-related technical problems included an unplugged in advance that the MRI room should be converted for surgical
or malpositioned coil (n¼5), a missing or non-functional conditions. Nevertheless, in 167 (88%) cases there was a delay
coil in the MRI room (n¼7), problems with positioning the in gaining access to the MRI room, this proportion ranging
head in the MRI scanner (difficulty placing the isocenter, from 62% to 100% depending on the time period (Fig. 1). The
n¼7), metal artifacts because of a surgical metal star holder average waiting time was 24 min (range 0 90 min).
or head clamp (n¼1 each), air artifacts, registration pro- Table 3 summarizes the time required to perform the
blems, MRI software bugs, air equilibration problems, a different steps of the ioMRI procedure. Transfer of the
non-MRI-adapted tube, and insufficient contrast agent (n¼ patient to the MRI room took a mean of 11 min (range
1 each). The artifacts occurring as a result of head position 2 40 min: patient #139, blocked transfer table needing system
(when repositioning was deemed unnecessary), metal or restart). The transfer time decreased, from a mean of 14 min
pneumencephaly did not prevent interpretation of the MRI. in the first group of 21 procedures to 8 min in the last cluster,
There were two unplugged or malpositioned coil problems paralleling the incidence of problems with the transfer table.
in the first three months, but only three in the subsequent The scanning procedure itself required a mean of 31 min
36 months. This problem was solved by re-plugging or (11 76 min) and the transfer back to the OR took a mean of
repositioning the coil and usually caused a significant delay 8 min (range 2 40 min: patient #27, problem with transfer
of about 15 min (5 38 min). table). Hence, the ioMRI procedure, including both trans-
The problem of a missing coil or one that could not be used fers, was completed in a mean time of 51 min (46 57 min)
for technical reasons was solved by using another coil. In one across the whole study period; this time decreased from
case (patient #170), a body coil was used and, although de- 55 min in the first half of the study period (105 procedures,
creasing the signal-to-noise ratio, this allowed satisfactory MR 21 months) to 47 min in the second half (84 procedures,
imaging. These problems occurred mostly in the last 6 months 19 months). Finally, the whole process, including waiting
with five incidents (1/4.2 ioMRI) and special attention is being time for access, took a mean time of 75 min (27 140 min),
paid to try and prevent this problem. The other MRI related which decreased slightly from 77 min during the first half of
problems occurred evenly though the 39-month period. the study period (105 procedures, 21 months), to 70.5 min in
There were also delays related to anesthetic issues (n¼7), the second half (84 procedures, 19 months).
with a mean duration of 13 min (2 20 min). These problems
included difficulties with monitoring material (n¼3), the
respirator (n¼2), SPO2, and one unknown (n¼1 each),
Discussion
and occurred with a relatively stable frequency of one case
every 5.71 months. Each anesthetic problem was successful-
ly handled and there were no related adverse events. Our data show that performing ioMRI at 3.0 T is a challenging
Waiting for access to the MRI room. Because the MRI procedure but remains feasible and safe. To our knowledge,
equipment is also used for routine diagnostic and research no untoward events related to the magnetic field strength have
activity, neurosurgeons always warn the MRI staff about 1 h ever been reported in the literature. Our experience shows
Fig. 1 Percentage of cases when

OR had to wait for access to
ioMRI and average waiting time
in function of team experience
Table 3 Average times in minutes (min max) for the different steps of the ioMRI process and their evolution over time
ioMRI Period (month) OR wait for MR Transfer to MR MRI duration Transfer to OR Total ioMRI
21 3 (Feb May 06) 17 (0 75) 14 (7 22) 28 (17 68) 9 (3 21) 72 (37 138)
21 4 (May Sept 06) 20 (0 44) 13 (7 35) 31 (17 60) 8 (2 40) 74 (53 120)
21 5 (Sept 06 Jan 07) 25 (5 75) 10 (4 15) 35 (16 55) 8 (3 15) 79 (45 135)
21 5 (Jan Jun 07) 21 (0 65) 11 (5 16) 37 (22 76) 8 (2 18) 78 (42 140)
21 4 (Jun Oct 07) 30 (10 60) 12 (5 25) 35 (11 70) 9 (3 25) 84 (55 124)
21 5 (Oct 07 Mar 08) 27 (0 55) 10 (5 15) 29 (17 60) 7 (2 25) 73 (35 130)
21 4 (Mar Jul 08) 19 (0 85) 11 (5 40) 29 (15 55) 8 (3 10) 67 (27 123)
21 4 (Jul Nov 08) 27 (0 90) 10 (3 30) 31 (15 52) 7 (2 25) 76 (36 138)
21 6 (Nov 08 May 09) 28 (0 60) 8 (2 10) 29 (15 55) 9 (2 23) 74 (34 120)
189 40 (Feb 06 May 09) 24 (0 90) 11 (2 40) 31 (11 76) 8 (2 40) 75 (27 140)
ioMRI intraoperative MRI (189 ioMRI with two in five surgical procedures), OR operating room
that even at ultrahigh fields, careful training can insure the The nature of the technical problems reported here
ioMRI procedure is safe. appears very similar to those observed in other series. Arte-
Every planned ioMRI procedure was completed success- facts on the MR images, because of ferromagnetic compo-
fully, enabling useful diagnostic information to be obtained. nents or pneumencephaly [5, 8, 9, 17 21], did not prevent
Technical problems occurred more frequently soon after the interpretation. Anesthetic equipment failure [12], difficulty
unit was inaugurated, with 15 technical problems during the positioning the head in the scanner [15, 22, 23], coil position
first 7 months of activity (32% of the 47 technical problems [10, 20], non-functional coils [17, 24], and internal MRI
experienced in more than 3 years). After this period, pro- system failure [15, 17, 19, 21] are difficulties that have
blems occurred occasionally, except for the non-functional been observed irrespective of the ioMRI system used. The
coil problems which occurred mainly in the last 6 months of number of technical problems reported here may seem
the study period. In these cases, the coil had to be changed, somewhat higher than those reported in literature, but in all
which did not prevent the ioMRI procedure being performed cases the problems were immediately resolved so that all
or limit interpretation of the images, but led to small delays. procedures were performed successfully. These problems
Our experience (Fig. 2) supports the notion that a learning should, therefore, be considered as minor issues and confirm
curve is observed in the initial period, after which specific the reliability and feasibility of performing ioMRI at 3.0 T.
technical requirements and equipment problems become Transport of the patients in our series required on average
more common [15 17]. 11 min to go and 8 min to return, with a range of 2 40 min
Intra-operative 3.0 T Magnetic Resonance Imaging Using a Dual-Independent Room 143
Fig. 2 Average times for ioMRI procedures in function of team experience
for both. These durations are in accordance with other series Conflict of interest statement We declare that we have no conflict of
where an independent surgical room was used [16, 19, 25, interest.
26], although some authors have reported a transfer time of
not longer than 3 min [23, 27, 28]. The average additional
References
time required to complete the whole ioMRI procedure was
75(27 140)min in our series. This duration is closely related
1. Black P, Moriarty T, Alexander E III et al (1997) Development and
to the ioMRI infrastructure and to the numbers and types of
implementation of intraoperative magnetic resonance imaging and
scanning sequences performed (e.g., whether or not an intra- its neurosurgical applications. Neurosurgery 41:831 845
venous contrast agent is used). ioMRI has been associated 2. Albayrak B, Samdani AF, Black PM (2004) Intra operative mag
with an increased duration of surgery of 1 4 h [3 12], but netic resonance imaging in neurosurgery. Acta Neurochir 146:
543 557
this could be decreased to 10 35 min [23, 27 30] with some 3. Archer DP, McTaggart Cowan RA, Falkenstein RJ et al (2001)
compromises in surgical conditions and imaging protocol. Intraoperative mobile magnetic resonance imaging for craniotomy
Another potential source of delay is the time needed to lengthens the procedure but does not increase morbidity. Can J
access the ioMRI room, which has not, to our knowledge, Anaesth 49:420 426
4. Black P, Alexander E III, Martin C et al (1999) Craniotomy for
been reported in other publications on this subject. Despite
tumor treatment in an intraoperative magnetic resonance imaging
an effort to schedule the time when the MRI scanner would unit. Neurosurgery 45:423 433
be needed, a large number of procedures (88%) were still 5. Knauth M, Wirtz CR, Tronnier VM et al (1999) Intraoperative MR
delayed by having to wait for access to the MRI room. The imaging increases the extent of tumor resection in patients with
high grade gliomas. AJNR Am J Neuroradiol 20:1642 1646
average waiting time overall was 24 (0 90)min. 6. Kremer P, Tronnier V, Steiner HH et al (2006) Intraoperative MRI
The dual independent twin room 3.0 T. ioMRI system for interventional neurosurgical procedures and tumor resection
offers several advantages as it provides MR images with control in children. Childs Nerv Syst 22:674 678
quality as good as a diagnostic scanner; it does not interfere 7. Schulder M (2008) Intracranial surgery with a compact, low field
strength magnetic resonance imager. Top Magn Reson Imaging 19
with the regular operating room in terms of surgical access, (4):179 189
patient positioning and instrumentation; it includes an accu- 8. Schulder M, Carmel PW (2003) Intraoperative magnetic resonance
rate navigation tool; and it enables alternative use of the imaging: impact on brain tumor surgery. Cancer Control 10:
magnet so that, when not needed by the neurosurgeon, it 115 124
9. Schulder M, Liang D, Carmel PW (2001) Cranial surgery naviga
can be used for routine diagnostic and research activity. tion aided by a compact intraoperative magnetic resonance imager.
Disadvantages include the need to move the patient during J Neurosurg 94:936 945
craniotomy and sometimes a wait for access to the MRI 10. Schulder M, Catrambone J, Carmel PW (2005) Intraoperative
room. Nevertheless, our large experience over more than 3 magnetic resonance imaging at 0.12 T: is it enough? Neurosurg
Clin N Am 16:143 154
years confirms that ioMRI at 3.0 T is a safe and feasible 11. Schulder M, Salas S, Brimacombe M et al (2006) Cranial surgery
technique with only minor problems. with an expanded compact intraoperative magnetic resonance
imager. Technical note. J Neurosurg 104:611 617
12. Siomin V, Barnett G (1998) Intraoperative imaging in glioblasto 21. Wirtz CR, Knauth M, Stamov M et al (2002) Clinical impact of
ma resection. Cancer J 9:91 98 intraoperative magnetic resonance imaging on central nervous
13. Schwartz RB, Hsu L, Wong T et al (1999) Intraoperative MR system neoplasia. Tech Neurosurg 7:326 331
imaging guidance for intracranial neurosurgery: experience with 22. Ntoukas V, Krishnan R, Seifert V (2008) The new generation
the first 200 cases. Radiology 211:477 488 Polestar N20 for conventional neurosurgical operating rooms: a
14. Jankovski A, Francotte F, Vaz G et al (2008) Intraoperative mag preliminary report. Neurosurgery 62(3 Suppl 1):82 89
netic resonance imaging at 3 T using a dual independent operating 23. Pamir MN, Ozduman K, Dinçer A et al (2009) First intraoperative,
room magnetic resonance imaging suite: development, feasibility, shared resource, ultrahigh field 3 Tesla magnetic resonance imag
safety, and preliminary experience. Neurosurgery 63(3):412 424, ing system and its application in low grade glioma resection.
discussion 424 426 J Neurosurg [Epub ahead of print]
15. Senft C, Seifert V, Hermann E et al (2008) Usefulness of intra 24. Nimsky C, Ganslandt O, von Keller B et al (2004) Intraoperative
operative ultra low field magnetic resonance imaging in glioma high field strength MR imaging: implementation and experience in
surgery. Neurosurgery 63(4 Suppl 2):257 266, discussion 266 267 200 patients. Radiology 233:67 78
16. Steinmeier R, Fahlbusch R, Ganslandt O et al (1998) Intraoperative 25. Nimsky C, Ganslandt O, Fahlbusch R (2002) From intraoperative
magnetic resonance imaging with the magnetom open scanner: patient transport to surgery in the fringe field: intraoperative appli
concepts, neurosurgical indications, and procedures: a preliminary cation of magnetic resonance imaging using a 0.2 Tesla scanner
report. Neurosurgery 43(4):739 747, discussion 747 748 the Erlangen experience. Tech Neurosurg 7:265 273
17. Wirtz CR, Knauth M, Staubert A et al (2000) Clinical evaluation 26. Nimsky C, Ganslandt O, Gralla J et al (2003) Intraoperative low
and follow up results for intraoperative magnetic resonance ima field magnetic resonance imaging in pediatric neurosurgery.
ging in neurosurgery. Neurosurgery 46(5):1112 1120, discussion Pediatr Neurosurg 38:83 89
1120 1122 27. Pamir MN, Peker S, Ozek MM et al (2006) Intraoperative MR
18. Bohinsky RJ, Kokkino AK, Warnick RE et al (2001) Glioma imaging: preliminary results with 3 tesla MR system. Acta Neuro
resection in a shared resource magnetic resonance operating chir Suppl 98:97 100
room after optimal imageguided frameless stereotactic resection. 28. Ramina R, Coelho Neto M, Giacomelli A et al (2009) Optimizing
Neurosurgery 48:731 742 costs of intraoperative magnetic resonance imaging. A series of 29
19. Nimsky C, Ganslandt O, Tomandl B et al (2002) Low field mag glioma cases. Acta Neurochir (Wien) [Epub ahead of print]
netic resonance imaging for intraoperative use in neurosurgery: a 29. Iseki H, Muragaki Y, Nakamura R et al (2005) Intelligent operating
5 year experience. Eur Radiol 12:2690 2703 theater using intraoperative open MRI. Magn Reson Med Sci
20. Schneider JP, Trantakis C, Rubach M et al (2005) Intraoperative 4:129 136
MRI to guide the resection of primary supratentorial glioblastoma 30. Lewin JS, Nour SG, Meyers ML et al (2007) Intraoperative MRI
multiforme a quantitative radiological analysis. Neuroradiology with a rotating, tiltable surgical table: a time use study and clinical
47:489 500 results in 122 patients. AJR Am J Roentgenol 189:1096 1110
Multifunctional Surgical Suite (MFSS) with 3.0 T ioMRI: 17 Months
of Experience
Vladimı́r Beneš, David Netuka, Filip Kramář, Svatopluk Ostrý, and Tomáš Belšán
Abstract The 3 T ioMRI in Prague is composed of two Keywords Glioma Intraoperative imaging Magnetic
independent suites: the operating theatre and the 3 T MR resonance Neuronavigation Pituitary adenoma
suite, both of which can and do work independently. They
are connected by a double door and a special transportation
system. The whole operating table is moved on rails to
and from the MR gantry. Anaesthesiological equipment is Introduction
built from paramagnetic material, which is also moved
to and from the MR suite. The integral parts of the multi- In the past years the possibility of intraoperative imaging is a
functional surgical suite (MFSS) are the neuronavigation fast developing field of neurosurgery [1 6]. Various modali-
system, electrophysiological monitoring, surgical micro- ties and systems have been developed. In April 2008, the 3 T
scope with availability of indocyanin green angiography intraoperative MR suite was opened at the Central Military
and fluorescence-guided glioma resection technique and en- Hospital (CMH) in Prague. The system is jointly operated by
doscopy equipment. The operating theatre is equipped in a the Department of Neurosurgery, CMH and Charles Univer-
normal fashion with the exception of a head holder that is sity and Department of Radiology, CMH. In the following
paramagnetic. paragraphs the main features of the MFSS will be described
MR radiologist and MR assistants are alerted approxi- and surgical procedures in which ioMRI has been used will
mately 30 min before the requested intraoperative and out- be summarised.
patient service is interrupted to clean the MR suite. The
ioMRI takes 15 20 min and immediately after the door
closes the out patient activity is resumed.
Intraoperative MR was performed in 332 surgeries in the
first 17 months of operation. The most frequent indications Materials and Methods
were pituitary adenomas, followed by gliomas. Other indi-
cations were less frequent and included meningiomas, caver- Multifunctional Surgical Suite (MFSS) description and
nomas, aneurysms, epilepsy surgery, intramedullary lesions, workflow (Fig. 1).
non-pituitary sellar lesions, metastases and various other The 3 T ioMRI in Prague is composed of two independent
surgeries. In 332 cases no technical or medical complication suites: the operating theatre and the 3 T MR suite (3.0 T
connected with ioMRI was encountered. Signa HDx, General Electric). Both can and do work inde-
pendently. They are connected by a double door and a
special transportation system (Maquet Viwas). The whole
operating table is moved on the rails to and from the MR
gantry. Anesthesiological equipment (anaesthesiological
machine Aestiva5/MRI, Datex Ohmeda, vital signs monitor
IMM MRI, Datex Ohmeda) is built from paramagnetic
V. Beneš (*), D. Netuka, F. Kramář, and S. Ostrý material, which can be moved to and from the MR suite.
Department of Neurosurgery, Charles University, Central Military
Hospital, U vojenské nemocnice 1200, 16902 Prague 6, Czech Republic
The integral parts of the MFSS are the neuronavigation
e mail: vladimir.benes@uvn.cz system (VectorVision Sky, BrainLAB), electrophysiological
T. Belšán monitoring (Eclipse, Axon Systems), surgical microscope
Unit of Radiology, Central Military Hospital, Prague, Czech Republic with availability of indocyanin green angiography and
146 V. Beneš et al.
Fig. 1 Operative room with an

ioMRI setting
fluorescence-guided glioma resection technique (OPMI Table 1 Neurosurgical procedures performed in ioMRI OR in the first
Pentero, Zeiss) and endoscopy equipment (both Storz and 17 months
Wolf endoscope). The operating theatre is equipped in a Pituitary adenoma 122 Hypothalamic hamartoma 3
Glioma 112 Lymfoma 3
normal fashion with the exception of a head holder that is
Meningioma 23 Carotid endarterectomy 2
paramagnetic (Integra for general electric). Cavernoma 15 Orbital tumor 1
Thirty minutes before the requested intraoperative MR, Aneurysm 11 Tuberculoma 1
the radiologists are alerted and out-patient service is inter- Other sellar lesions 10 Choroid plexus carcinoma 1
rupted to clean the MR suite. The ioMRI takes 15 20 min Epileptosurgery 7 Cavernous sinus schwannoma 1
and immediately after the door closes the out patient activity Intramedullary lesions 6 Spinal meningioma 1
is resumed. Metastases 6 AVM 1
Cyst/Absces 5 Intracerebral hematoma 1
Over the past 17 months (April 2008 August 2009), 332
surgeries with single or multiple intraoperative imaging
were performed. In the same period the surgical theatre lack of adequate coils and the learning curve of both neuro-
was used for 300 other neurosurgical procedures of all types surgeons and radiologists.
in which ioMRI was not called into service. Some 15 20 out-
patient MR studies are performed daily in the MR suite.
Discussion
Procedures in Which ioMRI Was Performed The primary use of ioMRI is for all types of glioma, where
ioMRI helps in increasing radicality and decreases the risk
of surgery. The same is true for pituitary adenomas in which
Use of the ioMRI is always determined before surgery. In a certain number of surgeries for recurrences can be avoided.
general, we use ioMRI in all endoscopic endonasal surgeries These two areas are clear and well documented [7, 8] and the
and in all glioma surgeries. All other indications are individ- subject of other presentations of our group in this issue.
ual. Table 1 summarises the types of surgery where ioMRI
was performed.
Immediate Postoperative Imaging
Results The system very easily allows for immediate postoperative

imaging. This imaging is performed immediately after
In 332 cases no technical or medical complication connected the wound suture and dressing while the patient is still
with ioMRI was encountered. In all cases the quality of under general anaesthesia. The value of these images is the
imaging was sufficient for the surgeon despite that the availability of immediate information. Images provide the
quality varied because of technical adjustments, a temporary neurosurgeon with information about the brain condition and
Multifunctional Surgical Suite (MFSS) with 3.0 T ioMRI: 17 Months of Experience 147
shift after surgery and they improve the decision about because of the intraoperative brain shift, this may be rather
postoperative care (e.g., should the patient be ventilated or difficult and peroperative monitoring is more precise.
awakened)? The images also serve as the baseline for sub-
sequent postoperative imaging. An immediate postoperative
MR study can also have legal value. Intraoperative Assessment of Acute Ischaemia
DWI images clearly show the extent of ischaemia after, e.g.,

Intraoperative Navigation aneurysm clipping, which would allow for immediate clip
replacement. In AVM surgery inadvertent occlusion of the
en passage artery can be disclosed.
Even with the implementation of navigation systems, the
surgeon can encounter problems in finding the lesion (e.g.,
deep-seated small cavernoma). In such a case ioMRI pro-
vides an obvious solution. The lesion can easily be re-navi- Biopsies
gated using the intraoperative images. Navigation and
renavigation are also used not only for the lesion remnants The exact position of the biopsy needle can be ascertained [12].
but also in biopsies, endoscopic procedures, in multiple
cystic lesion targeting, etc.
Cystic and Multicystic Lesions
Intended Partial Resection The exact and immediate decrease in cyst volume can be
shown and subsequent movement and changes of additional
In certain cases of very large meningiomas and some other cysts in multicystic lesions accurately targeted [13, 14].
lesions the surgeon plans a partial/subtotal resection for a
variety of different reasons, including age or unfavourable
general health of a patient and fear of over decompression
Spinal Cord Tumours
after the total resection. It is rather difficult to assess the
extent of resection in, e.g., a large cribriform plane meningi-
oma. Usually, the surgeon overestimates the extent of resec- The use of ioMRI is the same for spinal cord tumours as for
tion. It is obvious that the precise extent of surgery required brain gliomas. However, the safe extent of resection is
can easily be assessed by ioMRI [9]. probably more closely monitored by SSEPs and MEPs.
Extradural Spinal Tumours

Skull Base Tumours
Control of the extent of resection and resection of any
The role of ioMRI is the same as for gliomas and pituitary
remnants can be enhanced by ioMRI.
adenomas, namely enhancing radicality whilst concomitantly
decreasing the risks of surgery. However, in these tumours
ioMRI can also show the distance to vital structures during
piecemeal resection. The shift of the brain structures is easily
shown and in skull base tumours intraoperative navigation is Spine Surgery
always precise and helpful (the skull base structures do not
move). ioMRI can be useful in both classical and endoscopic Mastronardi et al. studied the prognostic relevance of
approaches. the postoperative evolution of intramedullary spinal cord
changes in signal intensity on magnetic resonance imaging
after anterior decompression for cervical spondylotic mye-
lopathy [15]. Although we are not convinced whether ioMRI
Extent of Resection in Epilepsy Surgery has any real value in this kind of surgery, it should be noted
that it might be applied.
ioMRI can be employed to assess the extent of resection All these potential areas have not yet been sufficiently
in comparison with preoperative planning [10, 11]. However, addressed and thus the possible benefits of ioMRI will be
148 V. Beneš et al.
Fig. 2 Preoperative (a) and intraoperative (b) corticospinal tract tractography in case on convexity meningioma closed to eloquent area
demonstrated more clearly in the future. The use of tracto- 2. Fahlbusch R, Ganslandt O, Buchfelder M, Schott W, Nimsky C
graphies during surgery or the integration of functional MR (2001) Intraoperative magnetic resonance imaging during trans
sphenoidal surgery. J Neurosurg 95(3):381 390
would certainly be of interest. It is probable that software 3. Hall WA, Liu H, Martin AJ, Pozza CH, Maxwell RE, Truwit CL
adjustment of functional images to the brain shift shown on (2000) Safety, efficacy, and functionality of high field strength
peroperative images would facilitate the availability of a interventional magnetic resonance imaging for neurosurgery. Neu
sufficient position of eloquent areas on intraoperative rosurgery 46(3):632 641
4. Jankovski A, Raftopoulos C, Vaz G, Hermoye L, Cosnard G,
images. This multimodal imaging seems to be the most Francotte F et al (2007) Intra operative MR at 3T: short report.
challenging field today (Fig. 2). JBR BTR 90(4):249 251
We can expect the use of this technology to spread in the 5. Steinmeier R, Fahlbusch R, Ganslandt O, Nimsky C, Buchfelder M,
near future, similarly to the widespread use of CT in the Kaus M et al (1998) Intraoperative magnetic resonance imaging
with the magnetom open scanner: concepts, neurosurgical indica
1970s, MR in the 1980s and endovascular technologies, tions, and procedures: a preliminary report. Neurosurgery 43(4):
radiosurgery and navigation systems in the 1990s. Novel 739 747
and exciting possibilities of intraoperative imaging can be 6. Sutherland GR, Kaibara T, Louw D, Hoult DI, Tomanek B,
anticipated. Saunders J (1999) A mobile high field magnetic resonance system
for neurosurgery. J Neurosurg 91(5):804 813
7. Nimsky C, Ganslandt O, Von Keller B, Romstöck J, Fahlbusch R
(2004) Intraoperative high field strength MR imaging: imple
Conclusions mentation and experience in 200 patients. Radiology 233
(1):67 78
8. Nimsky C, von Keller B, Ganslandt O, Fahlbusch R (2006) Intra
operative high field magnetic resonance imaging in transsphenoi
ioMRI and other types of intraoperative imaging are modern
dal surgery of hormonally inactive pituitary macroadenomas.
technologies that enable the neurosurgeon to perform precise Neurosurgery 59(1):105 114
and safe procedures. Whilst intraoperative imaging is cur- 9. Adeolu AA, Sutherland GR (2006) Intraoperative magnetic reso
rently used most frequently for gliomas and pituitary adeno- nance imaging and meningioma surgery. West Afr J Med 25
mas, its potential use in many other areas of neurosurgery (3):174 178
10. Kelly JJ, Hader WJ, Myles ST, Sutherland GR (2005) Epilepsy
will be addressed and studied in future investigations. surgery with intraoperative MRI at 1.5 T. Neurosurg Clin N Am 16
(1):173 183
Conflict of interest statement We declare that we have no conflict of 11. Schwartz TH, Marks D, Pak J, Hill J, Mandelbaum DE, Holodny AI,
interest. Schulder M (2002) Standardization of amygdalohippocampectomy
with intraoperative magnetic resonance imaging: preliminary ex
perience. Epilepsia 43(4):430 436
References 12. Vitaz TW, Hushek SG, Shields CB, Moriarty TM (2002) Interven
tional MRI guided frameless stereotaxy in pediatric patients.
Stereotact Funct Neurosurg 79(3 4):182 190
1. Black PM, Moriarty T, Alexander E 3rd, Stieg P, Woodard EJ, 13. Vitaz TW, Hushek S, Shields CB, Moriarty T (2001) Changes in
Gleason PL et al (1997) Development and implementation of cyst volume following intraoperative MRI guided Ommaya reser
intraoperative magnetic resonance imaging and its neurosurgical voir placement for cystic craniopharyngioma. Pediatr Neurosurg
applications. Neurosurgery 41(4):831 842 35(5):230 234
Multifunctional Surgical Suite (MFSS) with 3.0 T ioMRI: 17 Months of Experience 149
14. Walker JB, Harkey HL, Buciuc R (2008) Percutaneous placement 15. Mastronardi L, Elsawaf A, Roperto R, Bozzao A, Caroli M,
of an external drain of the cisterna magna using interventional Ferrante M, Ferrante L (2007) Prognostic relevance of the post
magnetic resonance imaging in a patient with a persistent cere operative evolution of intramedullary spinal cord changes in signal
brospinal fluid fistula: technical case report. Neurosurgery 63(2): intensity on magnetic resonance imaging after anterior decompres
E375 sion for cervical spondylotic myelopathy. J Neurosurg Spine 7
(6):615 622
Intra-operative MRI at 3.0 Tesla: A Moveable Magnet
Michael J. Lang, Alexander D. Greer, and Garnette R. Sutherland
Abstract This paper presents the development and imple- strength, and operating room (OR) configuration [1 5]. Over
mentation of an intra-operative magnetic resonance imaging the past several years, adoption of ioMRI by the neurosurgi-
(ioMRI) program using a moveable 3.0 T magnet with a cal community has tended towards higher-field systems. In
large working aperture. 1999, results related to the development and integration
Methods: A previously established prototype 1.5 T ioMRI of an ioMRI system based on a moveable 1.5-tesla (1.5 T)
program based on a ceiling-mounted moveable magnet was magnet were reported [5, 6]. The system was designed as
upgraded to 3.0 T. The upgrade included a short, 1.73 m, a patient-focused environment, as, when not needed for
magnet with a large 70 cm working aperture (IMRIS, surgery, the magnet is moved to an adjacent room. Serendipi-
Winnipeg, Canada), whole-room radio-frequency shielding, tously, this configuration was found to facilitate sharing of the
and a fully functional MR-compatible operating room (OR) technology between surgery and other disciplines, such as
table. Between January and September 2009, 100 consecu- diagnostic imaging. The 1.5 T prototype system has since
tive patients were evaluated at 3.0 T. been upgraded to 3.0 T.
Results: The ioMRI upgrade maintained a patient-focused Signal-to-noise ratio increases linearly as a function of
environment. When not needed for surgery, the magnet was magnetic field strength [7]. This translates into higher image
moved to an adjacent room. A large aperture and streamlined resolution, decreased acquisition time, enhanced soft-tissue
OR table allowed freedom of patient positioning while maintain- contrast, and improved capability for functional MRI
ing access and visibility. Working at 3.0 T enabled application of (fMRI), diffusion tensor imaging (DTI), and multi-nuclear
advanced imaging sequences to the full spectrum of neurosurgi- MR-spectroscopy [8]. Enhanced image quality, together
cal pathology in the ioMRI environment. The use of ioMRI with advanced software design, has encouraged widespread
continues to show unsuspected residual tumor in up to 20% of acceptance of 3.0 T MRI. Several centers have successfully
cases. There were no adverse events or technical system failures. integrated 3.0 T MR systems into various ioMRI configura-
Conclusion: An ioMRI program based a 3.0 T moveable tions (Table 1). Recent advances in 3.0 T technology have
magnet is feasible. By moving the magnet, the system main- resulted in the development of a relatively short magnet with
tains a patient-focused surgical environment and the ability a working aperture of 70 cm, allowing enhanced patient
to share the technology between medical disciplines. visibility and access, respectively [9]. The present report
describes the integration of such a magnet, and its use in
Keywords 3.0-Tesla Intra-operative imaging Magnetic 100 consecutive neurosurgical cases.
resonance imaging Neurosurgery
Introduction Materials and Methods
Intra-operative magnetic resonance imaging (ioMRI) has Technology

been used in neurosurgery for over ten years. The techno-
logy has varied in respect to magnet design, magnetic field
Magnet and Gradients
M.J. Lang, A.D. Greer, and G.R. Sutherland (*)

The ioMRI system includes a ceiling-mounted moveable
Department of Clinical Neurosciences, University of Calgary, Calgary,
Alberta, Canada 3.0 T magnet (IMRIS, Winnipeg, Canada) capable of transfer
e mail: garnette@ucalgary.ca into and out of the OR, which measures 1.73 m in length, with
152 M.J. Lang et al.
Table 1 Technical specifications of installed 3.0T ioMRI systems

Institution Scanner OR configuration Aperture (cm) Gradient (mT/m) Slew Rate (mT/m/ms)
University of Minnesota Philips Single room 60 40 200
Acibadem University Siemens Two room 60 45 200
Université Catholique de Louvain Philips Two room 60 40 120
Barrow Neurological Institute General Electric Two room 60 50 150
University of Calgary Siemens Moving magnet 70 45 200
an internal diameter of 90 cm, and a weight of 8,000 kg, attenuation of at least 100 dB. The number of electrical
compared to the 5,100 kg 1.5 T system. The working aperture devices in the OR necessitates extensive use of wave guides,
is 70 cm with shielded gradients in place. The 5-gauss fringe filtering, and conversion of electrical signals to fiber optics,
field measures only 43.5 m2. The system provides a large where possible. A copper mesh-impregnated glass panel
homogenous field of view measuring 505050 cm at was installed between the MR control-room and the OR,
isocenter, maximum gradient strength of 45 mT/m, and allowing uninterrupted patient visualization during imaging.
slew rate of 200 mT/m/ms (Table 1). Wooden doors were constructed for the magnet storage
alcove to provide additional safety. The ioMRI suite was
fitted with acoustic tiling to minimize emanating noise.
Operating Table
Success of the ioMRI program at the University of Calgary

can, in part, be attributed to a unique MR-compatible OR Clinical Material
table [5, 6]. The table includes integrated 3- or 4-pin head-
holders and hydraulic motors that allow the full functionality Prospective study of 100 consecutive patients undergoing
of a standard neurosurgical OR table. Currently in its fifth 3.0 T ioMRI-assisted neurosurgical procedures was per-
generation, the table is wider (56 cm), has a thinner vertical formed from January to September 2009. With the exception
profile (14 cm), and now includes table side rails. The table of thoraco-lumbar spine and peripheral nerve disease, the
is also able to rotated 90 , increasing the separation from the patients represent the full spectrum of neurosurgical patho-
5-gauss line, which has allowed the safe introduction of logy (Table 2), although the majority underwent treatment
endoscopy and other non-MR-compatible technology into for intracranial neoplasia.
the ioMRI environment.
Results
RF Coils
RF coils are one of the most critical components of an MRI Technology

system, owing to their impact on signal-to-noise ratio.
Coil design for ioMRI must take into account head fixation There has not been a single technological system failure in
compatibility, field sterility, and maintenance of surgical the current series, compared to earlier experience with the
access. The RF system is based on a transmit body coil 1.5 T prototype, in which 34 out of 986 cases (3.4%) were
and a number of different receive coils, depending on the accompanied by technical failure delaying sequence acqui-
application. The RF train can have up to 102 elements with sition (Table 2). The system has been successfully used with
32 channels, with software capable of turning elements on patients placed in the supine, prone and lateral positions.
or off, depending on the field of interest. This RF In the present series, there has not been a single adverse
technology reduces total acquisition time while maintaining event. The heaviest patient, who was positioned supine,
image resolution via the application of parallel imaging weighed 111 kg.
techniques. A range of imaging sequences, including T1-weighted,
T2-weighted, FLAIR, DTI and MR-angiography (MRA),
were successfully acquired in the intra-operative setting
RF Shielding (Fig. 1). The use of a dedicated image processor has greatly
improved the speed at which tractography images can be
In upgrading from 1.5 to 3.0 T, the decision was made to displayed, allowing modification of surgical approach as
change from local to whole-room copper RF shielding, with suggested by white matter anatomy (Fig. 2).
Intra-operative MRI at 3.0 Tesla: A Moveable Magnet 153
Table 2 Intra operative MRI cases patients imaged at 1.5 T. A total of 170 imaging studies were
Pathologic category 1.5 T (n 986) 3.0 T (n 100) performed on the 100 patients using the 3.0 T ioMRI tech-
Tumors nology.
Glioma 368 39 For patients undergoing intracranial neoplasm resection,
Meningioma 102 16 those requiring additional resection following intra-opera-
Pituitary adenoma 71 9
tive imaging are shown in Table 3. Unsuspected residual
Metastasis 17 3
Vasculara 94 10
tumor rates are only reported for the large 1.5 T data set,
Epilepsy 157 7 though residual tumor continues to be observed in up to
Spine 38 2 20% of cases imaged at 3.0 T. Though not detailed here,
Otherb 105 14 ioMRI has demonstrated additional benefits such as accurate
Technology failure 34 0 craniotomy placement or modified surgical decision-making
a
AVM, cavernoma, aneurysm, EC IC bypass, micro vascular decom (Fig. 3). In addition to craniotomy for tumor resection, 3.0 T
pression
b ioMRI has been applied to uses such as confirmation of
Ependymoma, choroids plexus carcinoma, neuroblastoma, neuro
cytoma, schwannoma, craniopharyngioma, lymphoma, radiation necro spinal decompression during cervical spine procedures and
sis, hydrocephalus post-operative evaluation of arterial patency following an-
eurysm clipping.
Patients
Patients in this 3.0 T series had a mean age of 4319 years

Discussion
(range = 13 months 83 years), with seven of the cases in this
series performed on pediatric patients, and a male to female An ioMRI system based on a moveable 3.0 T magnet has been
ratio of 1.2:1. This compares to a mean age of 4218 (range ¼ successfully integrated into the operating room while main-
4 months 83 years) and a male to female ratio of 1.1:1 in 986 taining a patient-centered environment. A highly effective
Fig. 1 Imaging quality at 3.0 T: ioMRI images (a, coronal FLAIR; b, axial FLAIR) obtained from a patient with a 4th ventricle ependymoma.
Post resection images (c, DTI imposed on T1 images) with superimposed tractography demonstrate intact cerebellar white matter anatomy
following resection (arrow). Images of a sphenoid wing meningioma with dural tail (d, T2 coronal; e, T1 axial with gadolinium enhancement).
MR Angiography (f) demonstrates a small intra cranial aneurysm (arrow)
Fig. 2 Effect of pre operative

tractography on surgical
planning: Prior to resection of a
temporal lobe lesion associated
with intractable epilepsy, DTI
based tractography illustrates
displaced, yet intact, middle
temporal gyrus white matter
tracts (small arrows). Absence of
functional inferior white matter
tracts (large arrow) suggests a
sub temporal surgical approach
Table 3 Impact of ioMRI on intracranial neoplasm resection

consideration prior to introducing such technology into the
Tumor type Number of cases Unsuspected
OR environment.
residual tumor (%) In similar fashion, an MR-compatible OR table with full
1.5T 3.0T 1.5Tb functionality continues to be the focal point of a patient-
a
Glioma centered ioMRI system. Paired with a mobile magnet, this
Grade I II 212 9 27 (13%) system permits the use of standard surgical technique, ac-
Grade III 89 20 18 (20%) quisition of intra-operative imaging, and avoidance of
Grade IV 67 10 15 (22%)
unnecessary and potentially dangerous patient transfer.
Meningioma 102 16 2 (2%)
Pituitary adenoma 71 9 25 (35%)
Notably, as the magnet can be moved out of the operating
a
World Health Organization classification
theatre, the full table functionality is essential for OR use
b
Unsuspected tumor data is reported only for experience at 1.5 T due to during non-imaging cases.
the small 3.0 T sample size Three principles must become axiomatic prior to general
adoption of ioMRI technology. First, it should be possible to
apply ioMRI to the entire spectrum of neurosurgical disease
without negatively impacting standard operative technique.
3.0 T ioMRI program is dependent on the development of Secondly, given the environment of fiscal responsibility, it
multiple, interrelated technologies. Improvements in RF should be designed to maximize cost-efficiency. Third, the
shielding, image processing, and RF coil design have each operating room must remain a patient-centered environment.
played a role. However, expanded working aperture and OR The application of 3.0 T ioMRI to date has upheld these
table functionality remain key components of a broadly appli- principles to varying degrees. Some investigators employ a
cable ioMRI system. research-dedicated design incapable of efficient shared dia-
The University of Calgary currently possesses the only gnostic usage [11]. A two-room concept, in which patients are
3.0 T ioMRI system with a working aperture of 70 cm. An transferred from the OR into a separate imaging suite, was
arithmetic increase in working space corresponds to a multi- created in order to defray cost via shared use, and obviates
plicative increase in the number of procedures to which the need to purchase MR-compatible instruments [12, 13].
ioMRI can be applied. Only with such expanded application Such a layout necessitates transportation of the anesthetized
will ioMRI be adopted by the neurosurgical community at patient out of the OR for imaging, which could place the
large. With the availability of a working aperture of at least patient at increased risk. Regardless of OR configuration, all
70 cm, the ioMRI magnet no longer dictates crucial deci- other 3.0 T ioMRI systems currently in place are limited
sions such as patient positioning and approach. In this way, somewhat by a small working aperture. Architectural design
movable ioMRI technology will empower surgeons to deve- and technological development of the ioMRI program at the
lop novel treatments without impinging upon the current University of Calgary have sought to systematically address
standard of care or exposing patients to undue hazard to these principles.
obtain imaging. The 70 cm aperture makes all standard Experience with 3.0 T ioMRI in 100 cases has been
patient positions compatible with ioMRI, and expands the encouraging as to potential improved patient outcome. It is
use of other surgical adjuncts, such as robotics [10]. While generally accepted that gross total resection of intracranial
magnets with field strengths higher than 3.0 T have been neoplasm improves patient survival, and recent studies sug-
developed, working aperture size must remain a principal gest that high-field ioMRI can increase the rate of gross total
Intra-operative MRI at 3.0 Tesla: A Moveable Magnet 155
Fig. 3 Changing surgery:

Surgical planning (a and
b, coronal T2) ioMRI images
obtained from a patient
presenting with seizures,
associated with a small right
temporal lobe lesion (arrows).
Following resection of the
suspected lesion, frozen section
showed neurons and glia,
consistent with normal brain.
Intra dissection imaging (c and
d, coronal T2) showed complete
resection of the lesion, obviating
the need for additional biopsy.
Subsequent histopathology was
consistent with a ganglioglioma
resection [14]. This series demonstrates the utility of intra- in the event of unintended vascular occlusion. In addition,
operative imaging in neurosurgery, including detection and susceptibility artifact related to titanium clips limits eval-
resection of unsuspected residual lesions. Imaging prior to uation of the aneurismal neck [17]. Development of MR-
patient positioning optimizes craniotomy placement and invisible clips, based on ceramics, may make such an
surgical corridor orientation, and provides a unique educa- evaluation possible [18].
tional environment for surgical planning. High-resolution Significant challenges remain. Development of MR-com-
DTI-based white-matter tractography has been used patible anesthesia equipment has advanced significantly, but
extensively in the operating room for surgical planning and field exposure continues to inhibit reliable measurement of
intra-operative evaluation of tract preservation [15]. Further ST-segment changes within the bore of the magnet. It is for
research is required to firmly correlate pathologic results this reason that patients with significant cardiovascular
with intra-operative imaging. co-morbidity have been excluded from ioMRI-assisted pro-
Continuing development of 3.0 T ioMRI technology por- cedures. Some have proposed that barriers, real or perceived,
tends increasingly comprehensive neurological imaging. to the ease of ioMRI acquisition limit its practical use. In the
Incorporation of fMRI into pre-operative imaging may become current series, workflow disruption has been the largest
a useful tool in the preservation of eloquent cortex [16]. constraint on the number of studies performed, as surgery
Chemical shift also scales with increased field, which, is interrupted for up to 30 minutes by the image acquisition
may make MR-spectroscopy a useful tool for the intra- procedure. The ability to operate under real-time imaging
operative identification of pathologic tissue [7]. MR- would allow ioMRI to actively guide surgery, as opposed to
angiography may also develop significant intra-operative confirming the presumed completion of surgical goals.
utility in the setting of vascular disease. However, it has Ultimately, cost is the true limiting factor in the widespread
been limited to craniotomy placement (particularly for adoption of 3.0 T ioMRI. Indeed, the cost of high-field ioMRI
patients with cavernous angioma or AVM), as image is routinely cited as the motivation for low-field systems,
acquisition time complicates aneurysm clip re-application despite inferior image quality. Were cost not prohibitive,
there would be little question as to the superiority of 3.0 T operative monitoring with high field strength MR imaging initial
ioMRI. As demonstrated by the case series published to date, results. Radiology 215:221 228
4. Nimsky C, Ganslandt O, Tomandl B, Buchfelder M, Fahlbusch R
3.0 T ioMRI can be safely integrated into the OR. However, (2002) Low field magnetic resonance imaging for intraioerative
the substantial cost of this technology must first be justified. use in neurosurgery: a 5 year experience. Eur Radiol 12:2690 2703
It is beyond question that scans obtained at 3.0 T are the 5. Sutherland GR, Kaibara T, Louw D, Hoult DI, Tomanek B, Saun
highest quality and most widely applicable ioMRI images ders J (1999) A mobile high field magnetic resonance system for
neurosurgery. J Neurosurg 91:804 813
available today. Integration of this technology will continue 6. Kaibara T, Saunders JK, Sutherland GR (2000) Advances in mobile
to drive innovation in surgical technique and equipment. That intraoperative magnetic resonance imaging. Neurosurgery 47:137
said, modern healthcare places increasing emphasis on de- 138
monstrable clinical utility prior to the widespread adoption of 7. Lin W, An H, Chen Y, Nicholas P, Zhi GH, Gerig G, Gilmore J,
Bullitt E (2003) Practical consideration for 3T imaging. Magn
new technology. Initiation of a randomized controlled trial to Reson Imaging Clin N Am 11:615 639
assess the impact of this technology on patient outcomes and 8. Tanenbaum LN (2006) Clinical 3T MR imaging: mastering the
cost-benefit ratios is desirable and now feasible, given the challenges. Magn Reson Imaging Clin N Am 14:1 15
existing number of higher-field ioMRI units. Importantly, 9. Sadick M, Schock KB, Gretz N, Schoenberg SO, Michaely HJ
(2009) Morphologic and dynamic renal imaging with assessment
the disruption of traditional surgical workflow caused by of glomerular filtration rate in a pcy mouse model using a clinical
intra-operative imaging remains unresolved. Integration of 3.0 tesla scanner. Invest Radiol 44:469 475
MR-compatible robotics within the magnet bore is, however, 10. Sutherland GR, Latour I, Greer AD, Fielding T, Feil G, Newhook P
an attractive solution to this limitation, and would allow (2008) An image guided magnetic resonance compatible surgical
robot: rapid communication. Neurosurgery 62:286 293
surgery to truly take place within an image. 11. Truwit CL, Hall WA (2006) Intraoperative magnetic resonance
imgaging guided neurosurgery at 3 T. Neurosurgery 58:ONS
Conflicts of interest statement Dr. G. Sutherland and A.D. Greer 338 ONS 345
hold shares in IMRIS (Winnipeg, Canada). M.J. Lang declares no 12. Jankovski A, Francotte F, Vaz G, Fomekong E, Duprez T, Van
conflict of interest. Boven M, Docquier MA, Hermoye L, Cosnard G, Raftopoulos C
Acknowledgements This work was supported by grants from Alberta dual independent operating room magnetic resonance imaging
Advanced Education and Technology, Western Economic Diversifica suite: development, feasibility, safety, and preliminary experience.
tion, and the Calgary Health Trust. Neurosurgery 63:412 426
13. Pamir MN, Peker S, Özek MM, Dinçer A (2006) Intraoperative
MR imaging: preliminary results with 3 tesla MR system. Acta
Neurochir 98:97 100
14. Hatiboglu MA, Weinberg JS, Suki D, Rao G, Prabhu SS, Shah K,
References Jackson E, Sawaya R (2009) Impact of intraoperative high field
magnetic resonance imaging guidance on glioma surgery: a pro
spective volumetric analysis. Neurosurgery 64:1073 1081
1. Black PM, Moriarty T, Alexander E 3rd, Stieg P, Woodard EJ, 15. Field AS, Alexander AL (2004) Diffusion tensor imaging in cerebral
Gleason PL, Martin CH, Kikinis R, Schwartz RB, Jolesz FA (1997) tumor diagnosis and therapy. Top Magn Reson Imag 15:315 324
Development and implementation of intraoperative magnetic reso 16. Hall WA, Truwit CL (2006) 3 Tesla functional magnetic reso
nance imaging and its neurosurgical applications. Neurosurgery nance imaging guided tumor resection. Int J Comput Assist Radiol
41:831 842 Surg 1:223 230
2. Hadani M, Spiegelman R, Feldman Z, Berkenstadt H, Ram Z 17. Sutherland GR, Kaibara T, Wallace C, Tomanek B, Richter M
(2001) Novel, compact, intraoperative magnetic resonance imag (2002) Intraoperative assessment of aneurysm clipping using mag
ing guided system for conventional neurosurgical operating rooms. netic resonance angiography and diffusion weighted imaging:
Neurosurgery 48:799 809 technical case report. Neurosurgery 50:893 897
3. Martin AJ, Hall WA, Liu H, Pozza CH, Michel E, Casey SO, 18. Sutherland GR, Kelly JJ, Boehm DW, Klassen JB (2008) Ceramic
Maxwell RE, Truwit CL (2000) Brain tumor resection: intra aneurysm clips. Neurosurgery 62:ONS400 ONS405
One Year Experience with 3.0 T Intraoperative MRI
in Pituitary Surgery
David Netuka, Václav Masopust, Tomáš Belšán, Filip Kramář, and Vladimı́r Beneš
Abstract A multifunctional surgical suite with intraopera- Introduction

tive 3.0 T MRI (ioMRI) has been operating at the Central
Military Hospital, Prague since April 2008. Our experiences Intraoperative MRI enables the surgeon to depict the extent
over the past year and the effect of ioMRI on the extent of of the neurosurgical procedure intraoperatively. This tech-
pituitary adenoma resection are evaluated. nique has been developed over the past 15 years. Current
Eighty-six pituitary adenoma resections were performed development goes in two directions. Low-field MRI scan-
in 85 patients with ioMRI in the first year of the ioMRI ners represent one of the directions [1, 2]. Nowadays, these
service. Pituitary adenoma suprasellar extension was present systems are mobile and less expensive. Moreover, the qua-
in 60 cases, invasion into cavernous sinus in 49 cases, and lity of images obtained by these systems has been improved
retrosellar growth in one case. The surgical goal was set in the past years. Still, the quality is inferior to preoperative
before surgery: either a radical resection (49 cases) or a images. The crucial question is whether to use high quality
partial resection (37 cases). preoperative images and compare them with inferior quality
In the group of patients where a decision for a radical intraoperative images. High-field MRI scanners represent
resection was taken the results are as follows: ioMRI con- another branch of development [3 5]. These systems are
firmed radical resection in 69.4% of the cases; ioMRI dis- more expensive and less mobile but produce better quality
closed unexpected adenoma residuum and further resection images.
led to radical resection in 22.4%. Surgical suite with intraoperative 3.0 T MRI has been
In the group of patients where a decision for a partial operating at the Central Military Hospital, Prague since
resection was taken, the results are as follows: no further April 2008. Our experiences over the past year and the effect
resection was perfomed after ioMRI in 51.3% of the cases of ioMRI on the extent of pituitary adenoma resection are
and further resection was performed after ioMRI in 48.7% of evaluated. The operative room and MRI setting are the
the cases. subject of separate presentation from our group in this
ioMRI seems to be a valuable tool to increase the extent issue. Therefore, in this paper we will limit our discussion
of pituitary adenoma resection. to our experience and results from endoscopic endonasal
procedures.
Keywords Intraoperative imaging Magnetic resonance
Pituitary adenoma
The first endonasal procedure in our ioMRI operative room

was performed in April 18, 2008. All procedures were per-
formed using the bilateral endonasal approach. A MRI com-
D. Netuka (*), V. Masopust, F. Kramář, and V. Beneš patible head holder was applied after induction of
Department of Neurosurgery, Central Military Hospital, Charles
University, U vojenské nemocnice 1200, Prague, Czech Republic
anaesthesia. Next, the head of the patient is registered to a
e mail: david.netuka@uvn.cz frameless navigation system that is used obligatorily. The
T. Belšán patient is undraped after tumour resection and transferred to
Unit of Radiology, Central Military Hospital, Prague, Czech Republic a MRI scanner after tumour resection. New MRI data are
158 D. Netuka et al.
sent to the navigation system and combined with the pre- Results
operative navigation data. This enables us to use intra-
operative data without any need to re-register the head of In the group of patients where a decision for a radical
the patient. All MRI sequences are analysed for pituitary resection was taken the results indicated ioMRI confirmed
adenoma remnant or any kind of surgical complication. radical resection in 34 cases (69.4%), ioMRI disclosed ade-
Then the decision is taken either to continue the surgical noma residuum in 13 cases (26.5%) and further resection led
resection or to check solely the operative field in order to to radical resection in 11 cases (22.4%).
perform closure of the sella if needed. ioMRI was evaluated as radical resection in two cases
Our MRI protocol consists of preoperative 3.0 T MRI (4.1%). However, early postoperative MRI disclosed a small
scanning followed by ioMRI on the same scanner. Six chan- residuum of adenoma. The second surgery was performed in
nel intraoperative coils are used for ioMRI. ioMRI is per- one of these two cases and a residuum of adenoma was
formed both in T1-weighted images with and without resected. ioMRI led to increased radical resection by
gadolinium and in T2-weighted images. Further, MRI is 22.4% (before ioMRI 69.4%, after ioMRI 91.8%). These
performed on the first or the second postoperative day. All results are summarised in Table 1.
these images are used to evaluate the extent of surgical In the group of patients where a decision for a partial
resection. resection was taken the results showed that no further resec-
In the first year of ioMRI service 110 endonasal proce- tion was perfomed after ioMRI in 19 cases (51.3%) and
dures were performed. Non-pituitary adenoma lesions were further resection was performed after ioMRI in 18 cases
treated in 15 cases. This group consisted of skull base (48.7%) in order to achieve a more extensive but still partial
carcinomas in three cases, surgical revision for cerebrospi- resection (Table 2).
nal fluid (CSF) leakage (n=3), surgery for spontaneous CSF Mortality was 0% in this group of patients while neuro-
leakage (n=2), meningioma (n=2), dermoid cyst (n=2), logical morbidity was 1.1% (one patient suffered unilateral
craniopharyngioma (n=1) and revision because of postoper- amaurosis; in fact, this patient had temporal field deficit on
ative haematoma (n=1). Altogether, 95 pituitary adenoma the same eye preoperatively). CSF leakage requiring redo
cases were treated. No ioMRI was performed in nine cases. surgery was present in four cases (4.7%).
The reasons why no ioMRI was performed include CSF ioMRI was perfomed once per surgery in 70 cases, twice
leakage after pituitary adenoma resection (n=4), patient in 14 cases and three times in two cases. There was neither
with pacemaker (n=1), patient with ferromagnetic material technical failure in performing ioMRI nor a patient safety
in the skull (n=1), surgery abandoned for medical reasons issue.
(n=1), surgery abandoned because of infection of the sphe-
noid sinus (n=1), or the patient had no insurance coverage
(n=1). For the remaining 86 pituitary adenoma surgeries,
ioMRI was performed. Discussion
The goal of surgery was set before the actual surgery was
undertaken: either a radical resection (49 cases) or a partial
Several papers have focused on ioMRI in sellar region sur-
resection (37 cases). The decision for a partial resection was
gery in the past few years. In 1999, Martin et al. published
chosen if there was major cavernous sinus invasion (lateral
their findings in five patients with pituitary adenoma [6].
to internal carotid artery), previous repeated surgeries, pre-
vious gamma knife surgery, or parasellar growth. The basic
characteristics of patients selected for radical resection are Table 1 Results in the group of patients where radical resection was
intended
as follows: 48 patients, 49 surgeries, and 21 women, and 27
Number of cases 49 100%
men (mean age 51.9 years). There were ten microadenomas
ioMRI confirmed radical resection 34 69.4%
and 38 macroadenomas in this group. Suprasellar growth of Wrong ioMRI evaluation 2 4.1%
pituitary adenoma was observed in 27 cases and cavernous ioMRI disclosed adenoma residuum 13 26.5%
sinus invasion in 15 cases. The mean vertical diameter of Radical resection after ioMRI 11 22.4%
pituitary adenoma was 18.6 mm. Basic characteristics of
patients selected for partial resection are as follows: 37
patients, 37 surgeries, and 14 women and 23 men (mean
age of 59.1 years). All pituitary adenomas were macroade- Table 2 Results in the group of patients where partial resection was
nomas. Suprasellar growth was present in 33 patients, caver- intended
nous sinus invasion in 34 patients, retrosellar growth in one Number of cases 37 100%
patient and parasellar growth in two patients. The mean No further resection after ioMRI 19 51.3%
vertical diameter was 23.6 mm. Further resection after ioMRI 18 48.7%
One Year Experience with 3.0 T Intraoperative MRI in Pituitary Surgery 159
This group used 0.5 T ioMRI. In two cases unexpected References

pituitary adenoma residuum was disclosed and resected
after ioMRI. Seminal papers on this topic are from Erlangen, 1. Gerlach R, du Mesnil de Rochemont R, Gasser T, Marquardt G,
Reusch J, Imoehl L, Seifert V (2008) Feasibility of Polestar N20,
Germany. Nimsky et al. evaluated the effect of ioMRI in 200 an ultra low field intraoperative magnetic resonance imaging sys
cases [7]. This series consists of both pituitary adenoma and tem in resection control of pituitary macroadenomas: lessons
glioma, as well as rare cases where a 1.5 T ioMRI scanner learned from the first 40 cases. Neurosurgery 63(2):272 284
was used. Radical resection was achieved in 56.2% of the 2. Hadani M, Spiegelman R, Feldman Z, Berkenstadt H, Ram Z (2001)
Novel, compact, intraoperative magnetic resonance imaging guided
piruitary adenoma cases before ioMRI, whereas the propor- system for conventional neurosurgical operating rooms. Neurosur
tion of radical resection increased to 87.2% after ioMRI. gery 48(4):799 807
The same group evaluated the effect of ioMRI in 33 cases 3. Hall WA, Liu H, Martin AJ, Pozza CH, Maxwell RE, Truwit CL
of growth hormone secreting adenomas [8]. ioMRI led to (2000) Safety, efficacy, and functionality of high field strength
interventional magnetic resonance imaging for neurosurgery.
resection of the pituitary adenoma in five cases. Hormone Neurosurgery 46(3):632 641
normalization was achieved in three of these cases. 4. Jankovski A, Raftopoulos C, Vaz G, Hermoye L, Cosnard G,
All these papers, including several others [9 11], show Francotte F et al (2007) Intra operative MR at 3T: short report.
the value of ioMRI in pituitary surgery. Our work supports JBR BTR 90(4):249 251
5. Sutherland GR, Kaibara T, Louw D, Hoult DI, Tomanek B,
this finding as well. Saunders J (1999) A mobile high field magnetic resonance system
for neurosurgery. J Neurosurg 91(5):804 813
6. Martin CH, Schwartz R, Jolesz F, Black PM (1999) Transsphenoi
dal resection of pituitary adenomas in an intraoperative MR unit.
Pituitary 2(2):155 162
Conclusions 7. Nimsky C, Ganslandt O, Von Keller B, Romstöck J, Fahlbusch R
(2004) Intraoperative high field strength MR imaging: implemen
tation and experience in 200 patients. Radiology 233(1):67 78
ioMRI is a valuable tool in pituitary adenoma surgery. The 8. Fahlbusch R, Ganslandt O, Buchfelder M, Schott W, Nimsky C
rate of radical resection (radical resection intention group) (2001) Intraoperative magnetic resonance imaging during trans
and more extensive resections (partial resection intention sphenoidal surgery. J Neurosurg 95(3):381 390
group) are increased. We stress the importace of careful 9. Jones J, Ruge J (2007) Intraoperative magnetic resonance imaging
in pituitary macroadenoma surgery: an assessment of visual out
ioMRI evaluation. In our experiences we have not observed come. Neurosurg Focus 23(5):E12
any technical ioMRI setting failures or issues involving 10. Nimsky C, von Keller B, Ganslandt O, Fahlbusch R (2006) Intra
patient safety. Finally, no increase of morbidity was ob- operative high field magnetic resonance imaging in transsphenoi
served in our series. dal surgery of hormonally inactive pituitary macroadenomas.
Neurosurgery 59(1):105 114
11. Schwartz TH, Stieg PE, Anand VK (2006) Endoscopic transsphe
Conflict of interest statement We declare that we have no conflict of noidal pituitary surgery with intraoperative magnetic resonance
interest. imaging. Neurosurgery 58(1 Suppl):ONS44 51
Intraoperative CT and Radiography
Intraoperative Computed Tomography
J.C. Tonn, C. Schichor, O. Schnell, S. Zausinger, E. Uhl, D. Morhard, and M. Reiser
Abstract Intraoperative computed tomography (iCT) has Keywords Computed tomography (CT) CT angiography
gained increasing impact among modern neurosurgical Intraoperative CT Intraoperative imaging Spine surgery
techniques. Multislice CT with a sliding gantry in the OR
provides excellent diagnostic image quality in the visuali-
zation of vascular lesions as well as bony structures inclu-
ding skull base and spine. Due to short acquisition times Introduction
and a high spatial and temporal resolution, various moda-
lities such as iCT-angiography, iCT-cerebral perfusion and The development of neurosurgical techniques and the
the integration of intraoperative navigation with automatic demand for treating more complex lesions with the aim of
re-registration after scanning can be performed. This allows continuous reduction of surgery related morbidity has gene-
a variety of applications, e.g. intraoperative angiography, rated an increasing demand for sophisticated intraoperative
intraoperative cerebral perfusion studies, update of cerebral imaging modalities. Various technologies are available,
and spinal navigation, stereotactic procedures as well as albeit with different indications. Intraoperative ultrasound
resection control in tumour surgery. Its versatility promotes has proven to be a rather easy to use, straight forward
its use in a multidisciplinary setting. Radiation exposure is solution for the intraoperative localisation of deep-seated
comparable to standard CT systems outside the OR. For lesions like cavernomas, metastases or haemorrhages as
neurosurgical purposes, however, new hardware compo- well as, to a certain extent, for resection control of tumours.
nents (e.g. a radiolucent headholder system) had to be Being rather inexpensive with no additional costs of instal-
developed. Having a different range of applications com- lation, the use of ultrasound is dependent on the personal
pared to intraoperative MRI, it is an attractive modality for experience of the user. In addition, despite improved tech-
intraoperative imaging being comparatively easy to install nology, its resolution may be limited depending on the
and cost efficient. particular type of pathology [1, 2]. For the resection of
high grade gliomas, the use of tissue fluorescence after
administration of 5-aminolevulinic acid (5-ALA) has been
proven to be efficient to delineate tumour margins and resi-
dual tumour tissue. The rate of ‘‘radicality’’ in the micro-
J.C. Tonn (*) surgical resection of high grade gliomas has been improved
Neurochirurgische Klinik und Poliklinik, Ludwig Maximilians significantly hereby. However, this technology is only useful
Universität München, Munich, Germany for malignant gliomas, especially glioblastomas; moreover,
Neurosurgical Department, Klinikum Grosshadern, Marchioninistr. 15,
81377 Munich, Germany the drug itself has so far not yet been approved by health care
e mail: joerg.christian.tonn@med.uni muenchen.de authorities in all parts of the world [3].
C. Schichor, O. Schnell, and S. Zausinger Intraoperative imaging using magnetic resonance (MR)
Neurochirurgische Klinik und Poliklinik, Ludwig Maximilians scanners, either high field or low field, provide good visuali-
Universität München, Munich, Germany sation of soft tissue abnormalities like gliomas, pituitary ade-
E. Uhl nomas, and other tumours. However, this technology is very
Neurochirurgische Klinik, Landeskrankenhaus Klagenfurt, Klagenfurt,
Austria
expensive and implementation into a pre-existing operating
theatre demands a lot of resources. Moreover, the use of
D. Morhard and M. Reiser
Institut für klinische Radiologie, Ludwig Maximilians Universität intraoperative MRI is limited for cranial applications due to
München, Munich, Germany the geometry of the scanner in relation to the table system.
164 J.C. Tonn et al.
The surgical workflow has to be adapted to the work in a media in case of CT angiography, a motor injection pump
magnetic field with special demands for surgical instruments (Stellant MEDRAD Inc., Indianola, Pennsylvania, USA)
as well as anaesthesiological monitoring systems [4 6]. was employed.
Recently, computed tomography (CT) technology has During image acquisition, the gantry moves over the
made a tremendous progress. Multislice scanners (e.g. patient without necessity to adjust the position of ventilation
40-slice scanning systems) provide a very high image spatial systems or catheters for scanning.
resolution with very short imaging acquisition times [7]. CT imaging was performed with a collimation of 40
Software for image processing has generated tools to pro- 0.6 mm at 120 kV and 140 mAs with a rotation time of 1 s for
vide high quality 3-D CT-angiography, which has been CTA and 201.2 mm at 120 kV and 350 mAs for cranial
shown to provide images at a quality level superior to CT. The automated dose modulation software was used.
MR-angiography. Quantitative perfusion parameters like Multiplanar reconstructions (MPR) were calculated with a
cerebral blood volume, blood flow and time to peak may slice thickness of 1 3 mm in axial, sagittal and coronal
be displayed as colour coded maps. By virtue of electronic planes. Data could be imported into the frameless infrared
artefact suppression and use of adequate acquisition para- based neuronavigation system (Vector Vision Sky). For
meters, imaging of metallic implants in spine surgery can be re-registration, the gantry of the scanner is equipped with
achieved and image distorsion can be eliminated to a large fiducials equivalent to the fiducials at the head holder or the
extent so that implants are precisely depicted and correction spinal fiducial clamp in case of spinal surgery.
can be performed, if necessary.
For intraoperative applications, CT scanners with a wide
bore are superior to those with narrow opening enabling Techniques of Intraoperative Computed
better access to the patient. The longitudinal coverage has
Tomography Angiography (iCTA) and
also been enhanced so that the whole body of the patient
(from head to toe) can be scanned without repositioning. Perfusion Computed Tomography (PCT)
This allows intraoperative CT scanners to be used by many
surgical disciplines (e.g. ENT surgeons, vascular surgeons, For the application of contrast agent, the injector is con-
traumatologists, orthopaedic surgeons) in addition to neuro- nected to a central line or at least 10 gauge lumen peri-
surgery. Thus, such an installation can be used in a multidis- pheral venous catheter. Monitoring of the whole procedure
ciplinary scenario and thereby enhancing cost effectiveness. is secured by direct visual contact through a lead-glass
After the development of a setting with a scanner with a window and indirectly via monitor camera from a neigh-
sliding gantry on rails and integrating a neuronavigation boring room to get a visual control of all parts of the OR.
device for cranial and spinal applications, we analysed For iCTA a CT scout for planning of the scan range of the
the usefulness of this setting for spinal and neurovascular head and upper neck is acquired first. Then CTA is acquired
surgery. Moreover, the impact on workflow was examined. in caudo-cranial scan direction from C1 to the vertex. A user
Experience has been gained according to the visibility in modified bolus tracking technique (repeated sequential CT
spinal and vascular surgery as well as work flow analysis, scans every 1 s roughly at the level of the carotid bifurca-
which will be presented here [8, 9]. tions to monitor the contrast arrival at the cervical arteries,
scan is then started manually when contrast enhancement in
the arteries is visible) and a weight-adapted contrast agent
protocol is used for CTA. This was done to obtain high
Methods/Technology contrast attenuation values in the cerebral arteries and low
contrast enhancement overlay in the veins and sinus.
A 40-slice-CT scanner (Somatom Sensation Open Sliding CTA is then followed by PCT. The scan range is manually
Gantry, Siemens Healthcare, Forchheim, Germany) with a selected to avoid beam hardening artifacts starting 1 cm
sliding gantry and a diameter of 82 cm mounted on rails superior to the aneurysm clip and with a distance of at
within the floor of the OR was installed in a pre-existing least 1 cm to head clamps. Five second after injection of
operating room. As operating table a carbon table plate, 50 ml contrast agent (Imeron 300) at 7 ml/s followed by a
segmented to allow virtually all neurosurgical positionings, saline flush of 50 ml at 7 ml/s the PCT starts. Therefore
was used (Trumpf, Puchheim, Germany). The system was sequential scans with 24 mm slice thickness are acquired
adjusted to a ceiling mounted navigation system (Vector every second over a period of 40 s. A standard, vendor
Vision Sky, BrainLab, Feldkirchen, Germany). Invasive given PCT analysis software is used for perfusion analysis.
head fixation was performed with a radiolucent head clamp Color-coded parameter-maps of cerebral blood flow (CBF),
(Mayfield radiolucent skull clamp A-2002, Integra, Plains- cerebral blood volume (CBV) and time to peak (TTP) are
borough, New Jersey, USA). For application of contrast calculated.
Intraoperative Computed Tomography 165
Results
Work Flow
After final positioning of the patient according to the envi-

sioned procedure, a ‘‘safety check’’ was performed. Hereby
the gantry is moved over the patient in order to detect any
possible collision during the scanning procedure. A special
anti-collision system prevents any conflict in case the move-
ment of the gantry is obstructed. All positionings including
complex ones like park bench were feasible; the only excep-
tion for intraoperative scanning is the semi sitting position.
In case of spinal instrumentation, pre-operative imaging
was performed at that time with the data fed into the naviga-
tion system in order to allow intraoperative navigation.
Hereby, the real position of the spine during surgery is the
basis for navigation. This is especially valuable if luxation of
the spine has to be reduced in the OR (e.g. in case of cervical
instability) since any navigation on the basis of image sets
obtained prior to the positioning of the patient on the OR
table may be misleading.
The intraoperative examination time for vascular lesions
including 3-D CT-angio and a perfusion-CT was 12 min
including 3 min for additional draping /undraping of the
patient for the scanning procedure, 1:30 min for image
acquisition and 3:00 min for reconstruction of the data set.
For spinal instrumentation a mean time frame of 9 min was
needed for scanning and data acquisition/image evaluation
Fig. 1 Reconstructed intraoperative control CT after cervicothoracic
until resumption of surgery.
navigated ventrodorsal stabilization C5 Th8 due to metastatic infiltra
tion Th1 4 from mamma ca, kyphotic instability and compression of
the spinal cord
Radiation Exposure
could be reduced to zero from previously 4.4% in the pre-iCT
era (Fig.1). Especially in cervical spine screw placements and
The highest radiation exposure for intraoperative CT- in the cranial cervical junction, no major misplacements of
scanning was obtained in CT angiography (including CT screws occurred. There was no increase of infection rate or
perfusion studies). Here, the mean effective dose of CTA other procedure related morbidity compared to our pre-iCT
and CT perfusion together was 3.69 mSv. This is comparable series. Duration of surgery in the cranial cervical junction
to 3.6 mSv, which is the value typically required for a was 12721 min, thoracic spine stabilisation (eight screws)
4-vessel catheter angiogram and which does not include a 17038 min and for lumbar stabilisation (four screws)
perfusion study [8]. 100 24 min, all including the time for imaging [10].
Evaluation of Imaging Vascular Neurosurgery
With CT guided spinal navigation, a computed spatial accura- In a pilot series of neurovascular surgery, intraoperative
cy of 0.80.1 mm could be achieved. In a first series a total of CT angiography and intraoperative CT perfusion were per-
414 screws were analysed. Intraoperative CT could detect a formed. The image quality of CT angiograms was rated
minor misplacement (2 4 mm) in 16 screws (3.8% of all excellent by a radiologist (D.M.) in all 13 cases (Fig. 2).
screws) and a major misplacement (>4 mm) in four screws CT perfusion imaging was rated excellent or good in 10/11
(1.0%). All misplaced screws could be re-positioned during the cases, in one case artefacts resulted in major degradation
same surgery. Hereby, the necessity for screw revision surgery of the image quality. CT angiography and CT perfusion
166 J.C. Tonn et al.
clamps with pins made of artificial sapphire were used,

artefacts due to the pins may be met. Recently polymeric
composite pins proved to be best suited in terms of artefact
reduction with grip force comparable to standard titanium
pins [11].
In comparison to intraoperative MRI, iCT is definitely
superior in intraoperative spinal imaging. Moreover, iCT
provides better depiction of bony structures and enables to
acquire high quality CT angiography and cerebral perfusion
studies. Furthermore, CT is especially useful for the control
of catheter placement, e.g. in shunt surgery. MR angio-
graphy is severely limited in the depiction of vascular struc-
tures adjacent to aneurysm clips, which is clearly a major
drawback of MR angiography. As a recent development
for vascular imaging, intraoperative fluorescence angio-
graphy (ICG) has been shown to be extremely helpful [12].
Whether iCTA and CT perfusion might be complementary is
subject of a presently ongoing study. Compared to intra-
operative MRI, iCT has a reduced sensitivity in the detection
and delineation of low grade gliomas and small pituitary
Fig. 2 Intraoperative CTA after clipping of an aneurysm of the right adenomas.
medial cerebral artery. View from the left side, showing artifact free Intraoperative CT imaging is a versatile, cost efficient and
vascular anatomy of the media bifurcation in the close vicinity of the easy to handle easy to install technology, with the potential
clip (Black arrow). Right sided pterional craniotomy (white arrow)
of a multidisciplinary use especially in spine, skull base and
vascular surgery.
changed the surgical strategy in two cases, in which clip
repositioning, either due to residual portions of the aneurysm Conflict of interest statement We declare that we have no conflict of
or because of an inadvertent occlusion of a vessel, was interest.
corrected. The image quality of iCTA and the intraoperative
perfusion map were rated by the surgeon to be adequate for
intraoperative decision making [8]. References
1. Rasmussen IA Jr, Lindseth F, Rygh OM, Berntsen EM, Selbekk T,

Xu J, Nagelhus Hernes TA, Harg E, Håberg A, Unsgaard G (2007)
Discussion Functional neuronavigation combined with intra operative 3D
ultrasound: initial experiences during surgical resections close
to eloquent brain areas and future directions in automatic brain
The results of our studies indicate that intraoperative CT shift compensation of preoperative data. Acta Neurochir (Wien)
(iCT) is a promising method enhancing precision of neuro- 149:365 378
2. Roth J, Biyani N, Beni Adani L, Constantini S (2007) Real time
surgery and ameliorating the outcome of patients. State of neuronavigation with high quality 3D ultrasound SonoWand in
the art CT technology incorporating multislice (multi detec- pediatric neurosurgery. Pediatr Neurosurg 43:185 191
tor row) systems (MSCT) allow for very short acquisition 3. Stummer W, Pichlmeier U, Meinel T, Wiestler OD, Zanella F,
times and excellent image quality as well as intuitive multi- Reulen HJ, ALA Glioma Study Group (2006) Fluorescence
guided surgery with 5 aminolevulinic acid for resection of malig
planar and 3D representations. Moreover, MSCT enables to nant glioma: a randomised controlled multicentre phase III trial.
perform advanced applications, such as 3-D angiography, Lancet Oncol 7:392 401
cerebral perfusion imaging and the visualization of implants 4. Black PM, Moriarty T, Alexander E 3rd, Stieg P, Woodard EJ,
in spinal surgery. In contrast to MR-based intraoperative Gleason PL, Martin CH, Kikinis R, Schwartz RB, Jolesz FA (1997)
imaging, there is no need for dedicated surgical instruments. nance imaging and its neurosurgical applications. Neurosurgery
The system can easily be handled with no negative impact on 41:831 842, discussion 842 845
the surgical workflow. Radiation exposure in modern 5. Nimsky C, Ganslandt O, Buchfelder M, Fahlbusch R (2006)
MSCT-systems is acceptable as compared to conventional Intraoperative visualization for resection of gliomas: the role of
functional neuronavigation and intraoperative 1.5 T MRI. Neurol
fluoroscopy. The learning curve for the staff including Res 28:482 487
anaesthesiologists and scrub nurses as well as surgeons 6. Nimsky C, Ganslandt O, Hastreiter P, Wang R, Benner T,
was very steep. Even if newly developed radiolucent head Sorensen AG, Fahlbusch R (2005) Preoperative and intraoperative
Intraoperative Computed Tomography 167
diffusion tensor imaging based fiber tracking in glioma surgery. 10. Zausinger S, Scheder B, Uhl E, Heigl T, Morhard D, Tonn JC
Neurosurgery 56:130 137, discussion 138 (2009) Intraoperative computed tomography with integrated navi
7. Hundt W, Rust F, Stäbler A, Wolff H, Suess C, Reiser M (2005) gation system in spinal stabilizations. Spine 34(26):2919 2926
Dose reduction in multislice computed tomography. J Comput 11. Ardeshiri A, Radina C, Edlauer M, Ardeshiri A, Riepertinger A,
Assist Tomogr 29:140 147 Nerlich A, Tonn JC, Winkler PA (2009) Evaluation of new radio
8. Schichor C, Rachinger W, Morhard D, Zausinger S, Heigl TJ, lucent polymer headholder pins for use in intraoperative computed
Reiser M, Tonn JC (2009) Intraoperative computed tomography tomography. J Neurosurg 111(6):1168 1174
angiography with computed tomography perfusion imaging in 12. Raabe A, Nakaji P, Beck J, Kim LJ, Hsu FP, Kamerman JD,
vascular neurosurgery: feasibility of a new concept. J Neurosurg Seifert V, Spetzler RF (2005) Prospective evaluation of surgical
112(4):722 728 microscope integrated intraoperative near infrared indocyanine
9. Uhl E, Zausinger S, Morhard D, Heigl T, Scheder B, Rachinger W, green videoangiography during aneurysm surgery. J Neurosurg
Schichor C, Tonn JC (2009) Intraoperative computed tomography 103:982 989
with integrated navigation system in a multidisciplinary operating
suite. Neurosurgery 64:231 239, discussion 239 240
Intraoperative CT in Spine Surgery
Wolf-Ingo Steudel, Abdullah Nabhan, and Kaveh Shariat
Abstract Background: In spinal instrumentation the mis- Introduction

placement of screws, cages and rods may cause neurovas-
cular complications. Therefore a large variety of methods Thousands of spine operations are performed every year in
have been used in recent years to reduce such complications Germany with instrumentation. Nowadays, the general stan-
especially by navigation techniques and intraoperative three- dard is to perform intraoperative fluoroscopy as well as a
dimensional fluoroscopy. The aim of this study is to answer post-operative X-ray control to demonstrate the correct
the question: will intraoperative CT improve the efficiency placement of the implants.
of the treatment as well as the safety for the patient at the Spine surgery, including decompression and stabiliza-
spinal instrumentation? Specific questions were: are the tion, can be challenging due to the small geometry of impor-
implants placed correctly and has decompression been per- tant bony structures such as facets and pedicles, anatomical
formed sufficiently? variations (i.e. the course of the vertebral artery within the
Methods: This is a prospective study in 100 patients axis), and the close anatomical relationship of the bone to
mostly with degenerative diseases, tumours and trauma. 80 the spinal cord, roots and vessels [1].
patients were treated by spinal instrumentation. The use of spinal implants may cause neurovascular
A helical CT (Somatom Emotion 2003) was used, which damage. This despite various intraoperative imaging tech-
is firmly bound to the OR table by a track system. nique largely depends on the disease as well as on the
Results: 569 implants were used: 159 vertebra body localization. It seems to be more difficult in the upper cervi-
screws and plates, 88 cages, 154 pedicle screws, 73 facet cal, the thoracic spine, the thoracolumbal and lumbosacral
joint screws and 95 rods. region and in patients with idiopathic or degenerative scoli-
There was malpositioning in seven patients (8.75%). 18 osis [2, 3]. There are various intraoperative methods of
of 154 pedicle screws were misplaced, 2 of 88 cages, and 4 spinal image guidance such as preoperative CT-based navi-
of 73 facet joint screws, for a total of 24 (7.6%). gation systems, standard fluoroscopy-based, 3D fluoroscopy
Conclusions: Intraoperative CT is a useful tool to check and ultrasound methods [4 6]. According to the literature,
the correct position of the implants used, the extent of 3D-fluoroscopy-assisted screw insertion increases the safety
decompression and the realignment as early as possible. It of a variety of routine and complex spinal procedures [5, 7].
therefore reduces second operations. A postoperative CT is The authors present a new device for intraoperative imaging
no longer necessary. in spinal instrumentation, the CT-suite an operation theatre
with a mobile CT. It allows real-time three-dimensional
Keywords Intraoperative CT Intraoperative imaging imaging of the whole spine and direct control and visualiza-
Spinal instrumentation tion of the correct position of the implants, the extent of
decompression, especially in patients with degenerative dis-
eases and tumours, and the realignment. In this study we
refer exclusively to the correct positioning.
W. I. Steudel (*), A. Nabhan, and K. Shariat

Department of Neurosurgery, Universität des Saarlandes, Kirrberger
Straße, 66421 Homburg, Saar, Germany
e mail: ingo.steudel@uniklinikum saarland.de
170 W.-I. Steudel et al.
Patients and Methods penetration of the pedicles and in complete lateral pedicle
screw penetration or in malplacement of a cage.
CT-Suite
The CT-suite consists of a mobile CT system (SOMATOM

Emotion 2003, Siemens, Erlangen). The system comprises a Results
mobile gantry with a diameter of 70 cm and a scanning
length of 153 cm, a mobile carbon OR table (Alphamaquet This is a prospective study of 100 consecutive patients with
1150, Maquet, Rastatt) and a workstation beside the OR with different pathologies (Table 1). Among them are 20 patients
a visual and video connection. There is firm binding of the without spinal instrumentation. Only a CT-guided biopsy
helical CT to the OR table by a track system in the floor. was done in six patients with spinal tumour and only a
Slices can be done from 1 to 10 mm. There are different decompressive procedure in another four. Only a facet joint
processing options like multiplanar reconstructions and infiltration was performed in ten patients with a degenerative
angio sequences with visualization of patient’s vessels. An radiculopathy.
examination of the spine is possible with adequate position- The Somatotom was easy and safe to handle. Care had to
ing from the atlanto-occipital junction to the sacrum. The be taken to avoid exposure of the OR team to radiation due to
operating table, which includes a special holding device for scanning. The team went from the OR into the workstation
the head, is a special manufacture made out of carbon fibres room during the CT-examination. The intraoperative scan-
(Fig. 1). ning time amounted to 15 30 min. There was never a prob-
lem with sterility. The evaluation of the slices and correct
positioning of the implants takes about 5 min. Visualization
of the positioning of the implants was good on the images
obtained from all patients.
Patients 569 different implants were used in 80 patients: 88 cages,
159 vertebra body screws and plates, 154 pedicle screws, 73
This is a prospective study on 100 patients of different ages faced joint screws and 95 rods (Table 2). The vertebra body
and sexes undergoing spine surgery for a variety of indica- screws include four patients with dens screws; the facet joint
tions and therefore using different types of spine fixation screws include 8 Magerl screws.
devices. The patient was placed in a supine or prone posi- There was an incorrect positioning of the implants in 7 of
tion, depending on the type of surgical procedure. A CT the 80 patients: The malpositioning occurred in 18 pedicle
examination was performed after placement of the implants. screws, 4 faced joint screws, twice in vertebra body replace-
The patient is always covered by a sterile plastic drape ment cages (Table 3).It was observed in two patients with a
usually used to drape the microscope. The accuracy of thoracic tumour, two patients with lumbar degenerative
the position of the implants was checked according to the listhesis, two patients with severe spondylodiscitis at the
classification of Liljenqvist et al. 1997 [8]. Repositioning cervical and lumbar region and another patient with a com-
was immediately done followed by another CT in medial bination of degenerative and idiopathic listhesis.
Fig. 1 Technical equipment:

The operating room used for
surgical procedures in which CT
is planned to be used is equipped
with a Siemens Somatom
Emotion 2003 Helical CT
Intraoperative CT in Spine Surgery 171
Table 1 Diagnoses: Characteristics of 100 patients and their surgical procedures

Number of patients, Age (range) Sex Spinal instrumentations, Other procedures
N 100 M F N 80
Cervical myelopathy 16 41 86 10 6 16
Trauma 15 44 86 9 6 15
Radiculopathy 20 30 82 10 11 10 10a
Spinal tumor 16 37 80 12 4 6 6b
4c
Degenerative lumbar listhesis 16 45 75 8 8 16
Congenital lumbar listhesis 7 32 57 3 4 7
Spondylodiscitis 4 42 86 3 1 4
Malformations 4 19 65 2 2 4
Chronic polyarthritis 2 43, 77 2 2
a
Facet joint infiltration
b
Biopsy
c
Decompression
Table 2 Implants characteristics in 80 patients

N Cages Vertebra body Pedicle screws Cervical faced joint Rods Total
screws, plates and transpedicular
Diagnoses screws
Cervical myelopathy 16 29 80 22 16 147
Trauma 3Kyphoplasty 15 13 12a 18 20b 12 65
Spinal tumor 6 6 12 14 19 9 60
Degenerative lumbar listhesis 16 19 74 32 125
Congenital lumbar listhesis 7 7 14 14 35
Radiculopathy 10 10 46 56
Spondylodiscitis 4 2 2 30 8b 8 50
Malformations 4 1 2 4 12b 4 23
Chronic Polyarthritis 2 1 5 2b 7
Total 80 88 159 154 73 95 569
a
Including dens screwing four times
b
Including 1 case with Magerl screwing C1/C2
Table 3 Evaluations of the positioning of the implants by intra frequency of complications is given very differently in the
operative CT literature. The complications depend on the difficulty of the
Patients N 80 N 7 (8.75%) spine operation: firstly on the kind of illness and its
Type Correct positionings Incorrect
seriousness and secondly also on their localisation. There
Pedicel screws 154 18
are differences in the region of the occipito-cervical, the
Facet joint screws 73 4
Cages 88 2 thoraco-cervical, the cervical and the lumbo-sacral junction
Total 315 24 (7.6%) [9]. Furthermore, it has to be considered that the quality of
the bone plays an important role as well in spinal instrumen-
tation. A 14 55% misplacement rate for pedicle screws
using standard techniques has been reported [10, 11].
Discussion
Complication Rate
Intraoperative Imaging in Spinal
Various types of spinal operations have relatively different Instrumentation
rates of complications, which may include damage to the
spinal cord as well as the roots of the nerves, injury to vessels It goes without saying that a wide range of treatments to
and nearby organs, and hardware failure. Complications complement conventional fluoroscopy has been developed
occur mostly through malpositioning of the implants. The over the years to improve and lower the complication rate.
Fig. 2 Intraoperative CT Scan

(left), orientation misplaced upper
right screw. Immediate revision,
control CT (right)
Fig. 3 Intra 3D CT (left) three dimensional CT (left) of a patient harbouring a sacral neurinoma. Tumour removal was done including lumbar
pelvic fixation (right). Note the different inclination and orientation of the screws within the spine and the pelvis
Fig. 4 Intraoperative CT Scan of a patient having received an occipital cervical fixation down to C3. Note the different orientation of the screws
within the occiput. The reconstruction of the position of the implants is on the right
Intraoperative CT in Spine Surgery 173
These, such as spinal navigation or 3D fluoroscopy, aim to Conflict of interest statement We declare that we have no conflict of
directly control the implanted material. 3D spinal image interest.
guidance significantly reduces the complications and break-
age rate and improves safe and accurate placement of spinal
instrumentation [11 22]. Using computer-navigated pedicle
screw insertion in the lumbar spine, a postoperative CT References
control no longer seems to be necessary (Bostelmann et al.
2005). Other treatments are ultrasound-methods and neuro- 1. Deinsberger R, Regatschnig R, Ungersböck K (2005) Intraopera
monitoring techniques as well as the direct intra-operative tive evaluation of bone decompression in anterior cervical spine
surgery by three dimensional fluoroscopy. Eur Spine J 14(7):
control of the anatomic structures of the implants [23].
671 676
However, technical problems still exist. It goes without 2. Kim YJ, Lenke LG, Cheh G, Rew KD (2005) Evaluation of pedicle
saying that additional treatments are being tested in order screw placement in the deformed spoine using intraoperativ plain
to continue to increase the efficiency of the operative treat- radiographs: a comparison with computerized tomography. Spine
30(18):2084 2088
ment. For this, the intraoperative computer tomography is
3. Lehmann RA Jr, Sasso RC, Helgeson MD, Dmitriev AE, Gill NW,
particularly suitable. Rosner MR, Riew KD (2007) Accuracy of intraoperatives plain
radiographs to detect violations of intralaminar screws placed into
the C2 vertebrae: a reliability study. Spine 32(26):3036 3040
4. Baldauf J, Müller U, Fleck S, Hinz P, Chiriac A, Schroeder HW
(2008) The value of intraoperative three dimensional fluoroscopy
Radiation Exposure in anterior decompressive surgery of the cervical spine. Zentralbl
Neurochir 69:30 34
5. Holly LT (2006) Image guided spinal surgery. Int J Med Robot 2
Radiation exposure of the surgeon, operating room staff and (1):7 15
the patient is a concern when placing instrumentation with 6. Kantelhardt SR, Bock CH, Larseen J, Bockermann V, Schillinger W,
the aid of fluoroscopy [24]. This should also be considered Rohde V, Giese A (2009) Intraosseous ultrasound in the placement
of pedicle screws in the lumbar spine. Spine 34(4):400 407
in intraoperative CT imaging. But most patients are serially 7. Ito Y, Sugimoto Y, Tomioka M, Hasegawa Y, Nakago K, Yagata Y
CT scanned after surgery, so in most cases there will be no (2008) Clinical accuracy of 3D fluoroscopy assisted cervical pedi
additional radiation exposure. cle screw insertion. J Neurosurg Spine 9(5):450 453
8. Liljenqvist UR, Halm HFH, Link T (1997) Pedicle screw instru
mentation of the thoracic spine in idiopathic scoliosis. Spine 22
(19):2239 2245
9. Rath SA, Moszko S, Schäffner PM, Cantone G, Braun V, Richter HP,
Intraoperative CT Antoniadis G (2008) Accuracy of pedicle screw insertion in the
cervical spine for internal fixation using frameless stereotactic
guidance. J Neurosurg Spine 8(3):237 245
Intraoperative CT has the advantage of precisely verifying 10. Kotil K, Bilge T (2007) Accuracy of pedicle and mass screw
placement in the spine without using fluoroscopy: a prospective
the position of the implants and the quality of the surgical
clinical study. Spine 8(4):591 596
procedure. In addition, the extent of decompression of the 11. Nottmeier EW, Seemer W, Young PM (2009) Placement of thor
bony structure is nicely visualized. Unlike 3D-fluoroscopy, acolumbar pedicle screws using three dimensional image gui
soft tissue and vessels can be demonstrated during the same dance: experience in a large patient cohort. J Neurosurg Spine 10
(1):33 39
examination (Figs. 2 4).
12. Beck M, Moritz K, Gierer P, Gradl G, Harms C, Mittlmeier T
Intraoperative CT has the disadvantage of not being an (2009) Intraoperative control of pedicle screw position using three
online control device like spinal navigation. However, in dimensional fluoroscopy. A prospective study in thoracolumbar
contrast to spinal navigation, which has to be considered to fractures. Z Orthop Unfall 147(1):37 42
13. Bostelmann R, Benini A (2005) Computer navigated pedicle screw
be virtual, intraoperative CT in spinal surgery does verify
insertion in the lumbar spine. Oper Orthop Traumatol 17(2):
that the surgical goals have been accomplished. In addition, 178 194
it is especially useful when implants have to be placed in 14. Darabi K, Reisch R, Müller Forell W (1998) Intraoperative com
very different orientations, as for example within the cranio- puted tomography in neurosurgery. In: Lemke HU et al (eds)
Computed assisted radiology and surgery. Elsevier Science,
cervical, cervicothoracic or spino-pelvic junction.
New York, pp 605 508
In addition to the value of this technique for intraopera- 15. Ebmeier K, Giest K, Kalff R (2003) Intraoperative computerized
tive imaging, economical advantages should also be men- tomography for improved accuracy of spinal navigation in pedicle
tioned. Surgical revision procedures because of incomplete screw placement of the thoracic spine. Acta Neurochir 85:105 113
16. Haberland N, Ebmeier K, Grunewald JP, Hliscs R, Kalff R (2000)
bony decompression or malpositioning of cages, screws and
Incorporation of intraoperatives computerized tomography in a
plates can be avoided. Postoperative CT examinations are no newly developed spinal navigation technique. Comput Aided
longer necessary to control implant positioning. Surg 5(1):18 27
17. Maier B, Zheng G, Ploss C, Zhang X, Welle K, Nolte LP, Marzi I methods for cervical pedicle screw placement and review of the
(2007) A CT free, intra operative planning and navigation system literature. Eur Spine J 17(4):564 575
for mnimally invasive anterior spinal surgera an accuracy study. 22. Yukawa Y, Kato F, Yoshihara H, Yanase M, Ito K (2006) Cervival
Comput Aided Surg 12(4):233 241 pedicle screw fixation in 100 cases of unstable cervical injuries:
18. Nottmeier EW, Foy AB (2008) Placement of C2 laminar screws pedicle axis views obtained using fluoroscopy. J Neurosurg Spine 5
using three dimensional fluoroscopy based image guidance. Eur (6):488 493
Spine J 17(4):610 615 23. Magit DP, Hilibrand AS, Kirk J, Rechtine G, Albert TJ, Vaccaro AR,
19. Ondra SL, Marzouk S, Ganju A, Morrison T, Koski T (2006) Simpson AK, Grauer JN (2007) Questionnaire study of neuro
Safety and efficacy of C2 pedicle screws placed with anatomic monitoring availability and usage for spine surgery. J Spinal
and lateral C arm guidance. Spine 31(9):263 267 Disord Tech 20(4):282 289
20. Papadopoulos EC, Girardi FP, Sama A, Sandhu HS, Cammisa FP Jr 24. Nardone E, Chen J, Maggio W, Nauta H (1996) The value of
(2005) Accuracy of single time, multilevel registration in image intraoperatives ultrasonography in cervical corpectomy: assess
guided spinal surgery. Spine J 5(3):263 267 ment by postoperative computed tomography. Neurosurgery
21. Reinhold M, Bach C, Audigé L, Bale R, Attal R, Blauth M, Magerl F 39:971 975
(2008) Comparison of two novel fluoroscopy based stereotactic
O-Arm Guided Balloon Kyphoplasty: Preliminary Experience
of 16 Consecutive Patients
Frédéric Schils
Abstract Balloon kyphoplasty is now widely used for the Introduction

treatment of vertebral compression fractures. Excellent pain
relief is achieved with cement injection, but the safety of the Vertebral compression fractures represent a major concern
procedure relys on excellent radiological exposure. The bal- amongst elderly patients, frequently leading to pain, dis-
loon kyphoplasty technique is usually performed using one ability and quality of life deterioration.
or two C-Arm devices to allow correct antero-posterior (AP) As the population continues to age, the challenge for the
and lateral view throughout the surgical procedure. By defi- treatment of osteoporosis and its complications will become
nition, this minimal invasive spine surgery is associated with increasingly prevalent [1 3].
radiation exposure for both the patient and the surgeon. In Conservative, non-surgical management of vertebral
our center, we recently moved from this way of proceeding compression fractures is actually challenged by surgical
to the use of an O-Arm image guidance system to perform minimally invasive procedures based upon cement augmen-
cement augmentation in vertebral fractures. tation (vertebroplasty and kyphoplasty). These procedures
To our knowledge, there is no clinical series describing may offer fast and sustained pain reduction, improved func-
the O-arm use in a balloon kyphoplasty procedure published tion while avoiding the kyphosis progression [4, 5]. Most of
in the scientific literature. We prospectively evaluate on 16 the studies in this field were conducted to demonstrate
consecutive patients, the feasibility of the O-Arm guided clinical benefit, vertebral deformity correction or risk of
kyphoplasty procedure with the original, usual tools, and cement leakage. To achieve these positive results, the pro-
we measured the fluoroscopy time and the X-ray exposure. cedure is performed under C-arm radiological exposure in
We didn’t experience any device related problem and most surgical teams. The C-arm is then manipulated from
demonstrated a significant reduction of X-ray exposure and AP to lateral view several times to provide secure pedicular
time of fluoroscopy. We believe that using this new intra- access and continuous fluoroscopy is required during
operative system, the overall time of surgery and fluorosco- cement injection. Although an impressive number of papers
py could still be reduced in a near future. about kyphoplasty is found in the literature, there is few data
about time of fluoroscopy during the surgery or dose of
Keywords Balloon kyphoplasty Minimal invasive spine X-ray exposure for the patient and his surgeon [6, 7]. Due
surgery O-arm intraoperative system Radiation exposure to the growing popularity of minimally invasive spinal
during surgery Vertebral compression fractures surgery, efforts to reduce X-ray exposure during the surgical
time to the patient and the surgeon may become a critical
goal in the future. We believe that new intraoperative
tools such as the O-arm system will help us to reach this
objective.
F. Schils
We started in February 2009 a prospective inclusion of new
Department of Neurosurgery, Clinique Saint Joseph, Liège 74, rue de
Hesbaye, B 4000 Liège, Belgium vertebral compression osteoporotic fractures admitted in our
e mail: frederic.schils@chc.be hospital for 3 months.
176 F. Schils
Patients were eligible for enrolment if they experienced (manufactured by Medtronic Spine LLC, Sunnyvale, CA,
one to three acute painful vertebral compression fractures USA), by a bilateral pedicular or extrapedicular approach.
from T5 to L5. Tumors, metastasis and myelomas vertebral All the procedures were conducted under general endo-
fractures were excluded from the study principally due to tracheal anaesthesia.
kyphoplasty reimbursement difficulties in theses indications Time of fluoroscopy by procedure and by level were
in Belgium. An hyperintense signal on MRI T2 and/or STIR precisely recorded as well as the radiation dose exposure.
sequences was required for the definition of recent compres- Overall operation time was also collected for each patient.
sion fractures.
Osteopenia evaluation through a bone mineral density
dexa scan was systematically used in all of our included
patients. Candidates for the study also had to have a mini- Population and O-Arm System
mal back pain score of 5 on a 0 10 visual analog scale
(VAS).
16 consecutive patients with 21 compression osteoporotic
Exclusions criteria were chronic fractures, rupture of any
fractures were involved in this prospective study and under-
kind of the posterior wall of the vertebra, infection condi-
went balloon kyphoplasty guided by the O-arm intraopera-
tions and any neurological deficit.
tive system.
O-Arm system is an intraoperative system based on a
conventional RX tube and a Flat Panel detector (4030
Varian). The system is used both in 2D mode, as a conven-
Procedure tional fluoroscopic system, and in 3D mode. The 3D mode is
particularly useful to evaluate cement distribution in the
Kyphoplasty was performed in all patients with classical vertebral body immediately after the procedure and to detect
instruments: introducers, inflatable bone tamps and poly- any cement leakage, in the vascular system, the spinal canal
methylmethacrylate bone cement and delivery devices or the intervertebral disc (Fig. 1).
Fig. 1 Illustration of a 3D O Arm acquisition at the end of the surgical procedure showing, in the three usual planes, the cement distribution in the
fractured vertebral body
O-Arm Guided Balloon Kyphoplasty: Preliminary Experience of 16 Consecutive Patients 177
Results offer several memorized positions (i.e. AP or lateral) reduc-

ing the manipulation required with a C-arm use and by the
16 consecutive patients with 21 vertebral compression frac- way the amount of unprofitable images and XR doses.
tures were evaluated during the three months of the study. The surgeon may always rely on the same AP and lateral
Mean patient age was 70 (range 62 88) and male/female predefined views during the overall procedure which allows
ratio was 19%/81%. All the 21 fractured levels were treated for higher security level and perfect visualisation of the
(12 patients with one level, three patients with two levels vertebral body in any plane at any time. In addition, the
and one patient with three levels). 53% of the levels were immediate 3D acquisition at the end of the procedure is
lumbar and 47% were thoracic. The mean surgical time (skin able to detect early, any complication and dispense the
to skin) for the procedure was 41 min (range 32 54) with need of a radiological postoperative control.
a mean fluoroscopy procedure time of 3.23 min (range These O-arm kyphoplasty procedures were conducted
2.68 5.04). The mean fluoroscopy time by level dropped to exactly the same way as C-arm procedures, without any
2.43 min. Mean irradiation dose by procedure was 247 mGy modification of the instruments and without any additional
and mean irradiation dose by level was 192.5 mGy. All recurrent cost. We observed a mean fluoroscopy time of
patients were addressed for a 3D scan performed at the end 3.23 min for a total procedure duration of 41 min. The
of the procedure. No cement leakage was found outside time of fluoroscopy by level was 2.43 min. These numbers
the vertebral body, avoiding any other postoperative radio- seems to be lower than those classically reported for opera-
logical control. tive or fluoroscopy time. For example, Izadpanah et al.
recently reported an operative time superior to 60 min with
computer navigation balloon kyphoplasty [14]. Boszczyk
et al. observed a mean fluoroscopy time of 3.8 min for single
Discussion level cases [7]. We are convinced that, due to a natural
learning curve, our values should be lowered in a very near
Several recent studies demonstrated the efficacity of kypho- future. New intraoperative devices such as the O-arm may
plasty to provide pain relief, improvement of quality of provide the surgeon easy and excellent radiological exposure
life, limitation in days of restricted activity and reduction of bone structures and may help to drastically reduce the
of analgesic use [8 11]. Recent publications also demon- X-ray exposure, the time of fluoroscopy and the duration of
strated the superiority of the kyphoplasty procedure over the the overall procedure with a highest degree of security.
non surgical, conservative treatment [5, 12]. Due to these
observations, and the growing prevalence of osteoporotic Conflict of interest statement We declare that we have no conflict of
interest.
vertebral compression fractures, the number of balloon
kyphoplasty procedures increased within the past years
[13]. Data concerning physician and patient X-ray exposure
during these minimally invasive procedures as well as pre- References
cise measurement of fluoroscopy time are still lacking.
To perform a kyphoplasty, the surgeon has to achieve 1. Borgstrom F, Zethraeus N, Johnell O et al (2006) Cost and quality
perfect visualisation of the pedicles in several planes and has of life associated withe osteoporosis related fractures in Sweden.
to be aware of possible cement leakage during the filling Osteoporos Int 17:637 650
2. Johnell O, Kanis JA (2006) An estimate of the worldwide preva
of the vertebral body. On an another hand, he has to try to
lence and disability associated with osteoporotic fractures. Osteo
reduce as much as possible the X-ray exposure for the poros Int 17:1726 1733
patient, the OR staff and of course himself. These two 3. Silverman SL, Minshall ME, Shen W, Harper KD, Xie S (2001)
goals may seem to be opposite. Several efforts to reduce The relationship of health related quality of life to prevalent and
incident vertebral fractures in postmenopausal women with osteo
the radiation time and dose are made through computer porosis: results from the Multiple Outcomes of Raloxifene Evalu
navigation kyphoplasty [14], and preventive protection mea- ation Study. Arthritis Rheum 44:2611 2619
sures such as protective coats, gloves and glasses. In com- 4. Garfin SR, Buckley RA, Ledlie J (2006) Balloon kyphoplasty for
puter navigation kyphoplasty, an additional incision has to be symptomatic vertebral body compression fractures result in rapid,
significant, and sustained improvement in back pain, function, and
performed in order to attach the reference clamp, and the
quality of life for elderly patients. Spine 31:2213 2220
cement filling of the vertebral body is still performed under 5. Grafe IA, Da Fonseca K, Hillmeier J et al (2005) reduction of pain
conventional fluoroscopy. and fracture incidence after kyphoplasty: 1 year outcomes of pro
Moreover, navigation represents also an additional recur- spective controlled trial of patients with primary osteoporosis.
Osteoporos Int 16:2005 2012
rent financial cost to the patient, or the hospital, due to 6. Seibert JA (2004) Vertebroplasty and kyphoplasty: Do fluoroscopy
numerous disposable single use tools. The O-Arm intra- operators know about radiation dose, and should they want to
operative system allows high spatial resolution and may know? Radiology 232:633 634
178 F. Schils
7. Boszcyk B, Bierschneider M, Panzer S et al (2004) Fluoroscopy 11. Ledlie J, Renfro M (2006) Kyphoplasty treatment of vertebral frac
radiation exposure of the kyphoplasty patient. Eur Spine J tures: 2 year outcomes show sustained benefits. Spine 31:57 64
15:347 355 12. Komp M, Ruetten S, Godolias G (2004) Minimally invasive the
8. Wardlaw D, Cummings S, Van Meirhaeghe J et al (2009) Efficacy rapy for functionally unstable osteoporotic vertebral fracture by
and safety of balloon kyphoplasty compared with non surgical care means of kyphoplasty: prospective comparative study of 19 surgi
for vertebral compression fracture (FREE): a randomised con cally and 17 conservatively treated patients. J Miner Stoffwechs 11
trolled trial. Lancet 373:1016 1024 (Suppl 1):13 15
9. Eck J, Nachtigall D, Humphreys S et al (2008) Comparison of 13. Tarantino U, Cannata G, Lecce D et al (2007) Incidence of fragility
vertebroplasty and balloon kyphoplasty for treatment of vertebral fractures. Aging Clin Exp Res 19 :7 11
compression fractures: a meta analysis of the literature. The Spine 14. Izadpanah K, Konrad G, Südkamp P (2009) Computer navigation
Journal 8:488 497 in balloon kyphoplasty reduces the intraoperative radiation expo
10. Bouza C, Lopez T, Magro et al (2006) Efficacy and safety of sure. Spine 34:1325 1329
balloon kyphoplasty in the treatment of vertebral compression
fractures: a systematic review. Eur Spine J
Intraoperative Ultrasonography
Intra-operative Imaging with 3D Ultrasound in Neurosurgery
Geirmund Unsgård, Ole Solheim, Frank Lindseth, and Tormod Selbekk
Abstract In recent years the quality of ultrasound (US) to the brain shift. There is an obvious need for intra-operative
imaging has improved considerably. The integration of imaging. One way to solve this is by intra-operative MRI.
three dimensional (3D) US with neuronavigation technology Another solution is 3D ultrasound. The main advantages with
has created an efficient and inexpensive tool for intra- 3D ultrasound compared to ioMRI are considerably lower cost
operative imaging in neurosurgery. Our experience is and higher flexibility. The disadvantage is that the surgeon
based on more than 900 operations with the intra-operative must learn how to use the technology to obtain optimal benefit.
3D ultrasound equipment SonoWand1 and some operations
with the research equipment Custux X. The technology has
been applied to improve surgery of intraparencymal brain System Description
tumours, but has also been found to be useful in a vide range
of other procedures, such as operations for cavernomas,
SonoWand1 is an integration of a high-end ultrasound scanner
skull base tumours, medulla lesions, arteriovenous malfor-
and a navigation system [1]. It can be used as an ordinary real-
mations (AVMs) and for endoscopy guidance. Compared to
time 2D ultrasound scanner or solely as a standard neuronavi-
intraoperative magnetic resonance imaging (ioMRI), 3D US
gation system based on pre-operative MRI. It also can
technology is advantageous in different ways: It is flexible
import MRI volumes with functional MRI and tractography.
and can be used in any operation theatre. There is no need for
However, it is the intra-operative 3D ultrasound capabilities
special instruments, and no need for radiologists or techni-
coupled with navigation technology that makes the system
cians. It adds very little extra time to the operation, and the
unique. Both 3D tissue and 3D angiography (power Doppler)
investment-costs are considerably lower than for ioMRI.
volumes can be acquired by tilting an ultrasound probe over the
area of interest. SonoWand1 enables simultaneous navigation
Keywords Arteriovenous malformation Cavernous
in both MR and US volumes. The volumes can be displayed
malformation Intraoperative imaging Intraoperative
either as axial/coronal/sagital slices or as oblique any plane
ultrasound Resection control Tumor
slices, steered by the pointer or different surgical equipment
(biopsy forceps, CUSA or endoscope) with a tracking frame
Introduction attached. Repeated 3D US uptakes during the operation, makes
it possible to navigate in an updated map that is very close to
the real time.
Neuronavigation systems have become standard tools for
planning of neurosurgery, but conventional systems have
limited value during the operation of brain tissue lesions due
Brain Shift
G. Unsgård (*) and O. Solheim
Department of Neurosurgery, St Olav University Hospital, Trondheim, In lesions with sharp definition of the lesion border we
Norway always found that the 3D ultrasound was correct (<2 mm
The Norwegian University of Science and Technology, Trondheim, off) while navigation based on pre-operative MRI was
Norway
e mail: geirmund.unsgard@ntnu.no
2 10 mm off from the true position. We also found conside-
rable inaccuracies in cases where there should be no brain
F. Lindseth and T. Selbekk
The Norwegian University of Science and Technology, SINTEF Health shift, for example in skull base tumours, reflecting the prob-
Research, Trondheim, Norway lem with accurate registration in spite of using fiducials.
182 G. Unsgård et al.
Image Quality Delineation of Tumours

The brain is very homogenous, thus enabling ultrasound Before using the 3D US images to guide operations, we felt
imaging with very few artefacts. Modern high-end scanners it was important to do a study to investigate the concurrence
with probes that are tuned optimally for brain scanning will between image delineation of the tumours and histopatho-
therefore depict all types of lesions with very good image logy. We found that there was a good concurrence between
quality (Fig. 1). We mostly use a 5 MHz (4 8 MHz) probe, the image interpretation at the biopsy site and the histo-
which gives optimal image quality at a distance of 2.5 6 cm pathology for biopsies sampled close to the image borders
from the probe. For superficial imaging a 12 MHz linear of the tumours [2].
probe is optimal. It produces superb image quality from the
first mm down to a depth of 4 cm (Fig. 3).
To be able to obtain good images throughout the opera-
tion, the patient must be positioned so that the planned Image Guided Resection
access to the lesion becomes vertical. In that way saline
will stay in the cavity during image acquisition. Our expe- By putting the optical tracking frame on the resection instru-
rience is that when this is considered during the planning, it ment (CUSA) it is possible to continuously follow the posi-
is almost always possible to obtain a setup where good tion of the CUSA during resection [3]. By looking partly in
quality images can be acquired throughout the operation. the microscope and partly on the navigation monitor the
The only situation where it is not possible to use ultrasound, surgeon can follow the position of the CUSA tip relative to
is when operating in sitting position. tumour borders and blood vessels. By making several 3D US
Fig. 1 Simultaneous display of preoperative MR data and intraoperative US data in the SonoWand system, during surgery of a low grade
astrocytoma (a), cavernoma (b), glioblastoma (c) and meningioma (d)
Intra-operative Imaging with 3D Ultrasound in Neurosurgery 183
uptakes throughout the operation the surgeon get a conti- Applications Where 3D US Has Improved
nuous update of his operating site compared to important the Surgical Technique
anatomical structures. This gives the surgeon information in
a much more dynamic and precise way than the microscope Low Grade Gliomas
interface in conventional neuronavigation systems.
(a) Very small low grade tumours. These low grade
tumours would be very difficult to operate without the
3D US technology (Fig. 3).
Resection Control (b) Large tumours in eloquent areas. Preoperative fMRI
and tractogrphy imported into the system is useful for
We have shown that residual tumour tissue missed during the planning of the operation. Intra-operative 3D US
microsurgery can be detected and removed with 3D US [4]. combined with the preoperative fMRI and tractogrphy
When all suspected tumour volume in the 3D US images has makes it possible to decide the limits of the resection in
been removed, there is usually no tumour left in the post- the 3D US volumes. We found that in cases where the low
operative MRI. It is also our experience that when we stop grade gliomas have ‘‘occupied’’ the expected site of the
the resection before the removal of all the high echo volumes pyramidal tract without causing neurological deficit, the
on 3D US due to eloquent area, we find tumour suspected tract was always pushed away by the tumour. Resection
areas on postoperative MRI. At the end of the operation of the tumour visible on 3D US did not create any perma-
there is sometimes artefacts in the wall of the resection nent neurological deficit. Our conclusion is that low grade
cavity (rim-effect). In those cases the images cannot be gliomas in eloquent area depicted by 3D US can be
trusted for the last 1 or 2 mm of resection. removed without risking permanent neurological deficit.
High Grade Gliomas

Impact on Surgery
3D US is useful to identify and resect all the extensions from
The use of intra-operative 3D US has influenced our opera- the high grade gliomas (Fig. 4). Even resection of high grade
tions in different ways: gliomas in eloquent areas can be done with very low morbidity
(a) It is easier to obtain a more radical resection of tumour and significant improvement of function [5].
tissue.
(b) It is much easier to do resections through very small
openings in the normal tissue. Thus we are less invasive Cavernomas
in the normal tissue with lower patient morbidity.
(c) We can work safer, faster and with more confidence
because the position of the resection instrument relative With conventional navigation it is possible to miss small
to an updated map of blood vessels, tumour borders and cavernomas due to brain shift and registration error. With 3D
normal tissue is continuously monitored. US all cavernomas are precisely imaged and localized. Even
very small brain stem cavernomas without lesion on the
surface are easy to find with 3D US.
Applications Where 2D US Has Improved

the Surgical Technique in Transsphenoidal Skull Base Tumours
Approaches
Resection with guided CUSA is useful to minimize the
A small side-looking high frequency linear array ultrasound trauma to the normal tissue, especially in meningiomas.
probe can be used to ensure orientation in the midline for the After an image guided subcapsular removal, the dissection
transsphenoidal approach, to identify important neurovascu- of the capsule is easier and less traumatic [6]
lar structures to be avoided during surgery and for identifi- This navigated subcapsular removal can proceed faster
cation of normal pituitary tissue and residual tumour and with more confidence because the surgeon will have 3D
(Fig. 2), The image resolution is far better than what can update of both the tumour borders and the blood vessels [7],
be achieved by current clinical MRI technology. and the surgeon knows exactly the progression of the operation.
Fig. 2 The use of ultrasound in pituitary surgery, with the probe and dimensions of the image overlaid on a MR image (a) and the intraoperative
US image showing residual tumour (indicated with *) (b), and the US probe with the side firing transducer array (c). In the middle of figure b is the
optic chiasm, and above is the communication artery
Biopsy Medulla Lesions
It is possible to make 3D US imaging through a15 mm burr 3D US has been successfully used for identification and
holes. Biopsies are sampled either free hand with calibrated biopsy of lesions in medulla [9], and for syrinxoperations.
biopsy forceps or with a guide.
AVM
Ventricle Catheter We have used 3D US angiography and stereoscopic display

of 3D MR angiography for early identification and clipping
3D US imaging of the side ventricle through 15 mm burr of feeders in AVMs [10]. In 25 patients 57 of 70 feeders
hole. Ventricle catheter with the stiff core is fixed to described in MR angiography could be identified and
a tracking device, calibrated and then the catheter act as a clipped in the beginning of the operation. This reduced
pointer. The core is removed when the catheter is in the the pressure in the AVM and made it much easier to dissect
ventricle. the AVM from the normal tissue.
Endoscopy Pros and Cons of 3D US
3D US is useful to guide the endoscope in complex multi- Navigated 3D US (SonoWand1) is not just another naviga-
cystic anatomy [8]. tion system. It is an intraoperative imaging system and a real
Intra-operative Imaging with 3D Ultrasound in Neurosurgery 185
Fig. 3 Preoperative MRI and

intraoperative US of a low grade
glioma, before (a), during (b), and
at the end of (c) resection of the
tumour. In figure b residual
tumour is seen to the left of the
cavity
competitor to intraoperative MRI. Even though there has • The active use of 3D US claims very little extra time
been no randomized study to look at efficiency of 3D US during surgery.
and ioMRI in obtaining radical tumour removal, we have • Investment costs of a SonoWand1 system is very low
indications to believe that 3D US gives similar intraopertive compared to any ioMRI.
information as ioMRI. What are the pros and cons when • When using 3D US, neurosurgeons have to be more
comparing 3D US to ioMRI? aware of the positioning of the patient.
• Neurosurgeons have to learn to make and interpret US
• 3D US is flexible; it can be used in any operation theater
images
• It is useful for nearly every type of neurosurgical operation
• The entire brain cannot be displayed in one 3D US volume,
• There is no need for special surgical instruments when
only the region of interest around the surgical approach.
using 3D US
• Neither is there any need for extra people (i.e. technicians There is a tendency among neurosurgeons to think that US is
and radiologists) a too simple technology. They are not aware of the huge
Fig. 4 A high grade glioma. Intraoperative 3D US overlayed the 3D MRI volume, before start of resection (a), during resection (b) and at the end
of resection (c). In the lower row is shown preoperative MRI (d) and postoperative MRI one day after operation (e)
improvement made to US in recent years. Some neurosurgeons ultrasound to delineate gliomas and metastases comparison of
are a little afraid of the challenge of learning a new technology. image interpretations with histopathology. Acta Neurochir 147
(12):1259 1269
It feels safer to leave the intraoperative imaging to the radiol- 3. Unsgaard G, Rygh OM, Selbekk T, Müller TB, Kolstad F, Lindseth F,
ogists. These feelings should not dominate our thinking. Due to Hernes TAN (2006) Intra operative 3D ultrasound in neurosurgery.
the rapidly increasing health care coast we all have an obliga- Acta Neurochir 148(3):235 253
tion to always go for the most cost-effective equipment. 4. Unsgård G, Ommedal S, Muller T, Gronningsaeter A, Hernes TAN
(2002) Neuronavigation by intraoperative 3D ultrasound, initial
experiences during brain tumor resections. Neurosurgery 50(4):
804 812
Conclusion 5. Gulati S, Berntsen EM, Solheim O, Kvistad KA, Håberg A,
Selbekk T, Torp SH, Unsgaard G (2009) Surgical resection of high
grade gliomas in eloquent regions guided by blood oxygenation level
Intra-operative 3D US is a flexible and inexpensive tool for dependent functional magnetic resonance imaging, diffusion tensor
minimizing the operation injury and increasing the efficien- tractography, and intraoperative navigated 3D ultrasound. Minim
cy and confidence of the surgeon during brain and medulla Invasive Neurosurg 52(1):17 24
6. Solheim O, Selbekk T, Linseth F, Unsgård G (2009) Navigated
operations. To have optimal benefit from this tool, the resection of giant intracranial meningiomas based on intraopera
surgeon has to learn some principles for the using of this tive 3D ultrasound. Acta Neurochir 151:1143 1151
technology in brain operations. 7. Rygh OM, Selbekk T, Lindseth F, Müller TB, Hernes TAN,
Unsgaard G (2006) Intraoperative navigated 3D ultrasound angio
Conflict of interest statement One of the authors (GU) holds 0.5% graphy in surgery. Surg Neurol 66:581 592
of the shares in SonoWand. 8. Rygh OM, Cappelen J, Selbekk T, Lindseth F, Hernes TANH,
Unsgård G (2006) Endoscopy guided by an intraoperative 3D
ultrasound based neuronavigation system. Minim Invasive Neuro
surg 49(1):1 9
References 9. Kolstad F, Rygh OM, Selbekk T, Unsgaard G, Nygaards OP
(2006) Three dimensional ultrasonography navigation in spinal
cord tumor surgery. Technical note. J Neurosurg Spine 5(3):
1. Gronningsaeter A, Kleven A, Ommedal S, Aarseth TE, Lie T, 264 270
Lindseth F, Langø T, Unsgård G (2000) SonoWand, an 10. Unsgaard G, Ommedal S, Rygh O, Lindseth F (2005) Operation
ultrasound based neuronavigation system. Neurosurgery 47(6): of arteriovenous malformations assisted by stereoscopic naviga
1373 1380 tion controlled display of preoperative magnetic resonance angio
2. Unsgaard G, Selbekk T, Müller TB, Ommedal S, Torp HS, Myhr G, graphy and intraoperative ultrasound angiography. Neurosurgery
Bang J, Nagelhus Hernes TA (2005) Ability of navigated 3D 56(2 Suppl):281 290
Intraoperative 3-Dimensional Ultrasound for Resection
Control During Brain Tumour Removal: Preliminary Results
of a Prospective Randomized Study
Veit Rohde and Volker A. Coenen
Abstract Introduction: The amount of resection is closely for complete tumour removal, 3-D ultrasound confirmed
related to survival in brain tumours. To enhance resection, complete tumour resection in three patients. In addition,
especially intraoperative magnetic resonance imaging 3-D ultrasound identified correctly one tumour remnant
(MRI) has been applied. The aim of this prospective, rando- in a patient randomized for complete tumour removal.
mized study was to test if intraoperative 3-D ultrasound Thus, the sensitivity for tumour remnant detection increased
likewise can be used for resection control. to 71% (five of seven patients) and that of confirmation
Methods: 16 patients, who underwent surgery for intraax- of complete tumour removal was 60 % (three of five
ial tumours in non-eloquent brain areas, were initially in- patients).
cluded into this prospective study. In two patients, the small Conclusion: The number of investigated patients is still to
size of the craniotomy hindered intraoperative ultrasound low to allow definite conclusions. However, the study results
imaging. In 14 patients, 3-D ultrasound images were suggest, that 3-D ultrasound is especially helpful for detec-
obtained before and after opening of the dura, during tumour tion of overseen brain tumour tissue.
removal, prior to evaluation by a blinded investigator for
identification of tumour remnants, and after dura closure. Keywords Glioma Intraoperative imaging Intraoperative
Seven patients were randomized to complete tumour remov- ultrasound Resection control
al according to the impression of the surgeon (group 1).
Seven patients were randomized to incomplete tumour
removal (tumour remnant <1 cm) (group 2); in these
patients, the neurosurgeon intentionally left a tumour
remnant prior to evaluation by the blinded investigator. Introduction
The tumour remnant was then removed. It was tested if
3-D ultrasound can correctly identify complete and incom- Several studies have indicated that time to recurrence in
plete tumour resection. All patients underwent early post- benign tumours and survival in malignant tumours is closely
operative MRI. related to the extent of resection [1]. With the aim to enhance
Results: In two patients (one each of the two groups) the the amount of resection, intraoperative magnetic resonance
image quality was too poor for a meaningful intraoperative imaging (MRI) and intraoperative computerized tomogra-
evaluation. In the six patients randomized for incomplete phy (CT) has been introduced into the neurosurgical routine
tumour removal, 3-D ultrasound correctly identified tumour [2 4]. Intraoperative MRI has the major disadvantages of
remnants in four patients (67%). In six patients randomized being expensive and requiring special non-ferromagnetic
equipment, if the operation is done in the environment of
the scanner, or special transport solutions which usually
V. Rohde (*)
Department of Neurosurgery, Georg August University Goettingen,
result in a reduced number of intraoperative re-investiga-
Robert Koch Strasse 40, 37075 Goettingen, Germany tions. The image quality of CT is inferior to that of MRI,
Department of Neurosurgery, Aachen University of Technology which even more holds true in the intraoperative situation,
(RWTH), Aachen, Germany and prevented the wide-spread use of intraoperative CT for
e mail: veit.rohde@med.uni goettingen.de
resection control.
V.A. Coenen
Department of Neurosurgery, Aachen University of Technology
The aim of this preliminary study was elucidate the
(RWTH), Aachen, Germany potential role of intraoperative 3-dimensional ultrasound
Department of Neurosurgery, University of Bonn, Bonn, Germany for intraoperative resection control.
188 V. Rohde and V.A. Coenen
Patients and Methods In additional two patients the image quality was too poor for a
meaningful intraoperative evaluation, leaving 12 patients for
A total number of 16 patients (nine men, seven women, mean evaluation. In the six patients randomized for incomplete
age 44 years) were included in this prospective randomized tumour removal, 3-D ultrasound correctly identified tumour
trial. The patients underwent microsurgery for cerebral metas- remnants in four patients (67%). In six patients randomized
tases (n¼8), malignant glioma (n¼6), and benign astrocytoma for complete tumour removal, 3-D ultrasound confirmed
WHO II (n¼2); tumours which would not have been complete- complete tumour resection in three patients. In this group,
ly resectable were excluded. Informed consent was obtained 3-D ultrasound identified correctly one tumour remnant which
from the patients and approval of the local ethic committee was was overseen by the surgeon and which was subsequently
given. Directly before surgery, imaging studies were performed removed after histopathological confirmation of tumour tissue.
with a 1.5 T MR scanner. All patients underwent an MRI Thus, the sensitivity of 3-D ultrasound for tumour remnant
anatomic navigation sequence preoperatively with fixed skin detection increased to 71% (five of seven patients) and the
fiducials. After administration of gadolinium, a T1-weighted sensitivity for complete tumour removal was 60% (three of
3-D fast field echo sequence was obtained in axial sections for five patients). In two patients, tumour remnants were
3-D volume rendering. The data set was transferred to the assumed on 3-D ultrasound, but not proven histologically.
SonoWand system (Mison AS, Trondheim, Norway), which Overall, the intraoperative situation was correctly predicted
is an ethernet-linked neuronavigational system with a high-end by ultrasound in 8 of 12 patients (67%).
ultrasound scanner. The system has the capability to trace
the navigated ultrasound probe, which allows acquiring 3-D
ultrasound volume sets which can be used again for neurona- Discussion
vigational purposes. The system displays preoperative MRI
data and the 3-D-ultrasound data simultaneously. Increased radicality of tumour resection is closely related to
Prior to surgery, the patient was randomized either to longer survival and longer progression-free survival in
complete tumour removal according to the neurosurgeons malignant brain tumours. The ill-defined border of intraaxial
impression (group 1) or to incomplete tumour removal brain tumours often prevents complete resection only with the
(group 2); in group 2, the neurosurgeon intentionally left a aid of the operating microscope. Thus, many attempts have
tumour remnant (<1 cm) prior to intraoperative ultrasound been made to improve the amount of resection. Stummer and
evaluation 4 (see below). Intraoperative 3-dimensional ultra- coworkers convincingly showed that the preoperative intake
sound images were obtained at five defined surgical steps: of 5-aminolevulinic acid and the intratumoural synthesis of
1-after trephination, before dura opening; 2-after dura open- fluorescent protoporphyrine IX, which ‘‘stains’’ the tumour,
ing; 3-at least once during tumour removal; 4-prior to evalu- allow reducing the percentage of incompletely resected
ation by the blinded investigator (Fig. 1); 5-after dura malignant gliomas significantly [5]. Even more effort was
closure. No information about the course of the operation made to bring MRI into the operating room for detection of
was given to the investigator, and the video screens display- overseen tumour. Intraoperative MRI has a high sensitivity
ing the intraoperative microscopic images of the resection for detection of tumour remnants, even if extravasation of
cavity were turned off during evaluation. The investigator contrast medium sometimes produces tumour-like artefacts
was allowed to see even repetitively all ultrasound images [6]. The high costs of intraoperative MRI and the necessity of
before making a final decision. It is noteworthy that the using non-ferromagnetic instruments if the operation is done
intentionally left tumour remnant was removed after the in the environment of the scanner hindered the wide-spread
ultrasound evaluation. It was investigated if intraoperative application of intraoperative MRI. A two-room solution with
3-D ultrasound allows differentiating between complete the need of a patient transport system is time-consuming and
tumour resection and tumour remnants. If the neurosurgeon limits the number of repetitive investigations.
has had the impression of complete tumour removal but the Intraoperative ultrasound has been proposed as a less
blinded investigator believed to have identified a tumour expensive and repetitively available alternative to MRI.
remnant on the intraoperative 3-D ultrasound image tissue Already in 1989 LeRoux and coworkers showed that
was harvested for histopathological investigation. the intraoperative ultrasound image correlates well with
histopathological findings, and Woydt et al. in 1996 con-
firmed these results [7, 8]. Both groups used 2-dimensional
Results ultrasound for identifying the tumour borders, and assumed
that ultrasound has the potential to enhance the amount of
In two patients standard craniotomy was not large enough for resection. Unsgaard and coworkers, using a 3-dimensional
obtaining 3-dimensional ultrasound images; sweeping the ultrasound in combination with neuronavigation, took tissue
ultrasound probe was hindered by the craniotomy margins. samples between 2 and 7 mm from the tumour border as
Intraoperative 3-Dimensional Ultrasound for Resection Control During Brain Tumour Removal 189
Fig. 1 Intraoperative 3 D ultrasound before (a) and after resection (b) of a large frontal anaplastic oligodendroglioma
seen on intraoperative 3-dimensional ultrasound images with intraoperative MR imaging might be more effective
and found a positive correlation of histology and ultrasound [11]. Our study is the first prospective randomized study
imaging in 74% low-grade gliomas, 83% astrocytomas WHO addressing the issue of resection control. As the investigator
grade III, 77% glioblastomas and 100% metastases [9]. who has to evaluate the ultrasound images was blinded with
Interestingly, studies focussing on resection control are respect to tumour remnant or total resection, an overall
rare: Lindner et al. used 3-dimensional ultrasound for resection correct prediction rate of 67% is acceptable, and in line
control in 23 patients with metastases, gliomas, cysts, lym- with the few published data [10]. Our study further indicates
phomas and meningiomas. Resection control was possible in that the strength of 3-D ultrasound is the identification of
78%, and the intraoperative ultrasound finding correlated tumour remnants; tumour remnants were correctly identified
with the post-operative MRI in 63.6% of the cases [10]. in 71%, which is remarkable because the tumour remnants
Gerganov et al. performed a prospective clinical trial and which have been intentionally left behind were less than
compared the reliability of intraoperative MR imaging 1 cm in diameter. The data of Gerganov et al. suggest that
and 2-dimensional ultrasound for tumour remnant detection. the detection rate would have been higher in cases of larger
Tumour remnants were seen in 21 of 26 patients by MR tumour remnants [11]. The correct identification of complete
imaging; with ultrasound two tumour remnants detectable by tumour removal was poorer with 60%; in two of three
MRI were missed, which indicates that resection control patients tiny blood clots in the resection cavity were mistaken
190 V. Rohde and V.A. Coenen
for tumour remnants. The same phenomenon already has 2. Albert FK, Forsting M, Sartor K, Adams HP, Kunze S (1994) Early
been described [11]. It can be assumed that the predictive postoperative magnetic resonance imaging after resection of
malignant glioma: objective evaluation of residual tumor and its
value would have been higher if the investigator would have influence on regrowth and prognosis. Neurosurgery 34:45 61
had a close knowledge of the resection cavity or would have 3. Uhl E, Zausinger S, Morhard D, Heigl T, Scheder B, Rachinger W,
performed the operation himself: Suspicious areas of poor Schichor C, Tonn JC (2009) Intraoperative computed tomography
delineation between tumour margin and normal brain tissue with integrated navigation system in a multidisciplinary operating
suite. Neurosurgery 64(5 Suppl 2):231 239, discussion 239 240
requiring a closer look up with ultrasound and blood clots 4. Wirtz CR, Knauth M, Staubert A, Bonsanto MM, Sartor K, Kunze S,
which can be mistaken for tumour on ultrasound imaging Tronnier VM (2000) Clinical evaluation and follow up results
already would have been known. for intraoperative magnetic resonance imaging in neurosurgery.
The low number of patients in our prospective study Neurosurgery 46:1112 1122
5. Stummer W, Pichlmeier U, Meinel T, Wiestler OD, Zanella F,
limits its value. Nonetheless, our data indicate that 3- Reulen HJ, ALA Glioma Study Group (2006) Fluorescence guided
dimensional ultrasound has the potential to control resection surgery with 5 aminolevulinic acid for resection of malignant
effectively during brain tumour surgery. The use of contrast- glioma: a randomised controlled multicentre phase III trial. Lancet
enhancing agents possibly would facilitate identification of Oncol 7:392 401
6. Knauth M, Aras N, Wirtz CR, Dörfler A, Engelhorn T, Sartor K
overseen tumour further [12]. (1999) Surgically induced intracranial contrast enhancement:
potential source of diagnostic error in intraoperative MR imaging.
AJNR Am J Neuroradiol 20:1547 1553
Conclusion 7. LeRoux PD, Berger MS, Ojemann GA, Wang K, Mack LA (1989)
Correlation of intraoperative ultrasound tumour volumes and mar
gins with preoperative computerized tomography scans. An intra
During brain tumour surgery, 3-dimensional ultrasound can operative method to enhance tumour resections. J Neurosurg
be used for resection control. Intraoperative ultrasound 71:691 698
seems to be most valuable for detection of tumour remnants 8. Woydt M, Krone A, Becker G, Schmidt K, Roggendorf W, Roosen K
and, thereby, offers the chance to enhance the extent of (1996) Correlation of intra operative ultrasound with histopatho
logical findings after tumour resection in supratentorial gliomas.
tumour removal. Proof of already completed total tumour Acta Neurochir 138:1391 1398
resection by 3-D ultrasound is not as reliable. Normal brain 9. Unsgaard G, Selbekk T, Brostrup Müller T, Ommedal S, Torp SH,
tissue altered by microsurgical resection can be mistaken as Myhr G, Bang J, Nagelhus Hernes TA (2005) Ability of navigated
tumour. To definitively define the value of 3-dimensional 3D ultrasound to delineate gliomas and metastases comparison
of image interpretations with histopathology. Acta Neurochir
ultrasound for resection control, a larger study is required. 147:1259 1269
10. Lindner D, Trantakis C, Renner C, Arnold S, Schmitgen A,
Conflict of interest statement We declare that we have no conflict of Schneider J, Meixensberger J (2006) Application of intraoperative
interest. 3D ultrasound during navigated tumor resection. Minim Invasive
Neurosurg 49:197 202
11. Gerganov VM, Samii A, Akbarian A, Stieglitz L, Samii M,
References Fahlbusch R (2009) Reliability of intraoperative high resolution
2D ultrasound as an alternative to high field strength MR imaging
for tumor resection control: a prospective comparative study.
1. Lacroix M, Abi Said D, Fourney DR, Gokaslan ZL, Shi W, J Neurosurg 111:512 519
DeMonte F, Lang FF, McCutcheon IE, Hassenbusch SJ, Holland E, 12. Kanno H, Ozawa Y, Sakata K, Sato H, Tanabe Y, Shimizu N,
Hess K, Michael C, Miller D, Sawaya R (2001) A multivariate Yamamoto I (2005) Intraoperative power Doppler ultrasonography
analysis of 416 patients with glioblastoma multiforme: prognosis, with a contrast enhancing agent for intracranial tumors. J Neuro
extent of resection, and survival. J Neurosurg 95:190 198 surg 102:295 301
Advantages and Limitations of Intraoperative 3D Ultrasound
in Neurosurgery. Technical note
Oliver Bozinov, Jan-Karl Burkhardt, Claudia Miranda Fischer, Ralf Alfons Kockro,
René-Ludwig Bernays, and Helmut Bertalanffy
Abstract Three-dimensional ultrasound (US) technology is alternative to MRI or CT [2 9]. However, neurosurgeons
supposed to help combat some of the orientation difficulties usually have little experience with the interpretation of
inherent to two-dimensional US. Contemporary navigation intraoperative (mainly oblique) US views and have instead
solutions combine reconstructed 3D US images with common been trained to interpret MRI images in the axial, coronal
navigation images and support orientation. New real-time and sagittal display [4, 8, 10, 11]. The growing spread of
3D US (without neuronavigation) is more time effective, ultrasound use has been influenced by significant improve-
but whether it further assists in orientation remains to ments in image quality over the past few years as well as
be determined. An integrated US system (IGSonic, the fact that probes have become small enough to fit into a
VectorVision2, BrainLAB, Munich Germany) and a non- minimally invasive craniotomy [9, 11, 12]. Our own group
integrated system with real-time 3D US (iU22, Philips, has described a two-platform model [13] and a one-platform
Bothell, USA) were recently compared in neurosurgical solution for the combination of MRI neuronavigation and
procedures in our group. The reconstructed navigation view US [11]. This led to increased routine use of this technolo-
was time-consuming, but images were displayed in familiar gy by several neurosurgeons previously unfamiliar with US
planes (e.g., axial, sagittal, coronal). Further potential applica- [11, 14]. The recent development of real-time 3D US
tions of US angiography and pure US navigation are possible. probes raises the question of whether one modality can
Real-time 3D images were displayed without the need for an replace the other. We present here our initial experience.
additional acquisition and reconstruction process, but spatial
orientation remained challenging in this preliminary testing
phase. Reconstructed 3D US navigation appears to be
superior with respect to spatial orientation, and the technique 3D Ultrasound-Assisted Image-Guided
can be combined with other imaging data. However, the Neurosurgery
potential of real-time 3D US imaging is promising.
Keywords Image-guided surgery Reconstructed 3D Technical Aspects

ultrasound Intraoperative sonography Neuronavigation
Real-time 3D ultrasound Neurosurgery The technical characteristics of the VectorVision2 (BrainLAB,
Heimstetten, Germany) system and the integrated US device
(IGSonic) have been reported previously in detail [11]. Brief-
Introduction ly, the precalibrated current IGSonic Probe 3000 (4 9 MHz
frequency) with multi-focus imaging is connected to the
VectorVision navigation station, and the infrared cameras
Recently it has become increasingly attractive to add
track the reflective marker spheres of a reference clamp
intraoperative imaging data to the surgical planning/proce-
attached directly to the probe. No registration of the US
dure [1], and ultrasound offers an inexpensive and quick
probe is necessary when using a one-platform solution. The
system provides real-time US information and overlays the
O. Bozinov (*), J. K. Burkhardt, C.M. Fischer, R.A. Kockro, ultrasonic images or puts them next to the corresponding
R. L. Bernays, and H. Bertalanffy
classical images of the navigation system. Acquisition of a
Department of Neurosurgery, University Hospital, Frauenklinikstrasse
10, 8091 Zürich, Switzerland 3D data set is possible with the most recent version of the
e mail: oliver.bozinov@usz.ch system.
192 O. Bozinov et al.
3D Ultrasound Acquisition and Display Ultrasound Angiography
After craniotomy, the first US investigation is performed once As previously reported by our group, intraoperative land-
the US probe is automatically registered (approx. 30 s). The marking of anatomical structures (e.g., vessels or tumour
draped touch screen allows alteration by the surgeon, who can remnants) by US is possible and may be fused to intraopera-
adjust the parameters for better visualisation and acquire tive 3D US reconstructions [10]. This might provide addi-
the 3D US dataset. The US probe is moved over the field tional useful intraoperative information for the surgeon in
of interest. Each real-time 2D US image is saved with the cases of complex glioma or vascular lesion [9, 10, 15, 16].
specific coordinates recognised by the navigation system Three-dimensional acquisition in the Doppler mode stores
through the US reference star. A maximum of 256 slices only colour-coded Doppler information, which in summary
were collected per dataset, and US data were reconstructed will reconstruct a 3D vessel superimposed on the navi-
based on spatial resolution. The acquisition time for an entire gation images of choice (Fig. 1b). Eventually, this leads
256-image set was 1 2 min, and the calculation time was to intraoperative non-invasive angiography. Our system
30 60 s each. Each 3D US dataset can be displayed in the did not provide any surface-rendering program, as CT or
usual planes of neuronavigation (axial, coronal and sagittal) MRI angiographies do. Other groups have demonstrated
(Fig. 1a, b). If a tumour border has been outlined preopera- enhanced image qualities by additional surface-rendering
tively (segmented image) in the MRI or CT dataset, it can be [9, 17]. Thus, more promising tools may be available in the
optionally superimposed onto the 3D US image (Fig. 1a, b). near future.
Fig. 1 The right side of the left navigation screenshot (a) of a right temporal cavernoma displays all three familiar planes (axial, coronal and
sagittal) of the 3D ultrasound reconstruction data corresponding to their preoperative navigated MRI data on the left side (highlighted by the small
circles in all six windows). The arrows point to the skull base below the cavernoma (hyperintensive in ultrasound reconstruction). The right
screenshot (b) includes the US angiography of the right internal carotid artery and middle cerebral artery (arrows) of a right temporal glioblastoma
case (outlined preoperatively). The navigation images of the preoperative MRI and reconstructed 3D ultrasound acquisition were also placed in
(b) as described above in (a)
Advantages and Limitations of Intraoperative 3D Ultrasound in Neurosurgery. Technical note 193
Real-Time 3D Ultrasound Imaging 3D Ultrasound Acquisition and Display

Technical Aspects
An acquisition procedure is not necessary, as the 3D data are
immediately available (live). In the ‘‘X-Plane’’ mode, two
The US probe X7-2 (2 7 MHz frequency) used to acquire US images are displayed (Fig. 2c). The left one always
real-time 3D images (iU22, Philips, Bothell, USA) is pre- shows the 2D US image according to the position of the
dominantly designed for paediatric cardiologists or gynae- probe (oblique). The right one displays the corresponding
cologists. The xMATRIX array technology utilises 2,400 live 2D image according to the chosen degree of turn (from
fully sampled elements for 360-degree focusing and steer- 0 to 360 ). For example, if one holds the US probe in an axial
ing. The array probe enables live xPlane imaging to acquire position, then a choice of a 90 or 270 turn will illustrate the
two full-resolution planes simultaneously from the same corresponding coronal (from the lateral approach) or sagittal
heartbeat or region of interest. The system’s multi-direction- (from the anterior approach) image of this area of interest
al beam steering provides unlimited planes in all directions, (Fig. 2c). The ‘‘Live 3D’’ mode displays all of the US
and live volume imaging allows the acquisition and render- information from the array probe at once in a cone. The
ing of full volume data at true real-time frame rates with width of the cone can be adjusted to the user’s needs.
unparalleled isovoxel resolution. The settings on the system The image is displayed online and changes immediately in
are not primarily ordered for neurosurgical procedures, response to all movements of the probe. In predominant
according to the company. However, there is increasing tissue areas (brain and tumour tissue with no cysts or ven-
corporate interest in further adapting parameters for brain tricles), visualisation of the field of interest is difficult; it is
and tumour visualisation, which is crucial for optimal US thus especially complex to demonstrate in a 2D picture
imaging. (Fig. 3). However, the overall advantage of this mode for
Fig. 2 Images from a 12

year old male show a cerebellar
low grade astrocytoma with
multiple cysts (a). The real time
3D ultrasound images (b) display
the solid lesion (L), cysts (C) and
tentorium (T) in two planes
simultaneously. The arrow
shows the degree of the right
ultrasound image compared
to the oblique left side. In this
case, the right image is a 270
turn of the left image (near axial)
and should represent a nearly
sagittal position. The ‘‘X Plane’’
mode was used for this image
Fig. 3 A cone picture is

displayed from the use of the
‘‘Live 3D’’ full volume mode
(same case as Fig. 2) with real
time 3D ultrasound (a). Basically,
all ultrasound data from the array
probe are shown in one 3D
animation. The width of the cone
can be changed according to the
user’s needs. The orientation is
very difficult and anatomical
structures not clearly presented.
Most of the tissue overshadows
the solid lesion (L), cysts (C)
and tentorium (T). The small
picture from the array probe
(b) demonstrates the linear
ultrasound in a 3D representation
(Copyright Philips, Bothell,
USA)
neurosurgeons still remains to be discovered. Gynaecolo- after filling the cavity with normal saline. After dural clo-
gists and pregnant mothers appreciate this ‘‘Live 3D’’ sure, images were mostly unacceptable (especially after 3D
mode, as it is often able to show the surface of the foetus reconstruction) due to remnant air bubbles. Resection con-
and sometimes even the face. Multiple sets of 3D US images trol is therefore recommended intraduraly and directly,
or videos can be stored and reactivated any time during the either with the navigated 2D mode or real-time 3D probe.
surgical procedure. The technical installation of both systems for surgery needs
no significant aid. However, this is different during surgery.
Navigated 3D US can be managed via a draped touchscreen
by the surgeon, whereas the real-time US machine has to be
Ultrasound Angiography managed by an unscrubbed person. It is possible to drape the
buttons and a small additional touch screen, but handling the
The usual Power Doppler and additional modes are available trackball through a drape is impossible.
with the real-time 3D US probe, but no advanced reconstruc-
tion for angiographies or navigation combination is included.
3D Orientation and Resolution

Comparative Advantages and Disadvantages
In all cases, a leading anatomical landmark (falx, tentorium,
ventricles, brainstem or skull base) close to the lesions was
Clinical Arrangements noted to allow for rapid orientation of the surgeon using
the real-time 3D US mode. Anatomical landmarks became
To facilitate optimal US imaging, nearly all patients are unnecessary when reconstructed 3D US images were pre-
positioned during surgery with a nearly vertical craniotomy sented next to their corresponding image slices of the pre-
corridor. In such a position, saline remains in the resection operative navigation sets (MR or CT) (Fig. 1a, b). The
cavity after the dura is opened for ultrasound imaging; the combination of both 3D data sets (US and MRI) in one picture
amount of air bubbles can be limited to reduce acoustic is also possible, but using the same greyscale for both images
shadows [9]. This technique was used to facilitate sono- makes visualisation difficult. Lindseth et al. [8] introduced
graphic resection control with both systems. In all cases, multimodal image fusion in US-based neuronavigation by
the chosen craniotomy was sufficient for the US probe; our improving the overview and interpretation. These authors
group performed no enlargement of the craniotomy for integrated preoperative MRI with intraoperative 3D US and
either US probe. Initial US acquisitions were performed added an accuracy study to this combination of two systems.
extraduraly. Repeated intraoperative image updating was Displaying the 3D data as a specifically oriented image plane
valuable and easily accomplished, even in inexperienced was somewhat supportive for the orientation. It was initially
hands. The best image acquisition was always obtained challenging to determine the degree that would actually
Advantages and Limitations of Intraoperative 3D Ultrasound in Neurosurgery. Technical note 195
become a true axial, coronal or sagittal slide (Fig. 2). The full a state of the art tool in the last two decades. Even in eco-
real-time 3D US cone image (‘‘Live 3D’’) provided no addi- nomically challenging countries, neuronavigation systems
tional help regarding orientation (Fig. 3). Unsgaard et al. [9] are increasingly purchased, however such devices are rarely
have published a benchmark review article regarding develo- used in practice because preoperative MRI is often too costly
pments in reconstructed 3D US. They have used the intra- for patients in these countries. Many neurosurgical groups
operative imaging system SonoWand (Mison A/S), which is a intend to obtain intraoperative imaging technology but often
high-end US platform with a supplementary navigation sys- struggle with the enormous costs of intra-operative MRI
tem [18]. This system has shortened the acquisition time to a imaging and furthermore the running costs should not be
very tiny delay of surgery. The resolution of its US probes is forgotten. An integrated US navigation system would offer
also superior to that of the IGSonic probe. The resolution a cheap intraoperative imaging alternative with familiar
of our real-time 3D probe is very good, but the image orientations (i.e., axial, coronal and sagittal) and basically
pre-settings could be improved. no running costs. Furthermore, this tool can also be used
without any expensive preoperative imaging modality (CT
or MRI) and therefore should be very interesting for eco-
nomically challenging countries as a true alternative to
Comparison via One Lesion Entity intraoperative MRI or CT. The real-time 3D US system itself
(Cavernoma) (with additional transducers) is certainly a high-end, modern
intraoperative imaging device with minimal running costs.
Cavernomas provide a favourable group of lesions for
US-guided resection [19]. Their significant hyperechogenity
(Fig. 1a) is most likely due to the concentrated hemosiderin in Outlook and Future
and around the lesion, as seen on haemorrhages in US images.
After 3D acquisition, neuronavigation was sometimes
changed completely to the US mode without using any further Further development of image quality for one-platform solu-
preoperative MRI data. Vessels close to the cavernoma (like tions would be highly appreciated, but those improvements
the often-seen draining vein) can be visualised and land- should not change the convenient manageability, handiness
marked nicely with the Color Doppler mode [10]. Real-time or plug-and-play usage. Most interesting will be further
3D ultrasound was also able to localise and demonstrate such technical developments in real-time 3D US sonography
lesions. No disturbing shadows were experienced due to per- with better image settings, real time segmentation features,
fect 3D visualisation. However, navigation guidance for fur- higher frequency, multi focus imaging and larger probe
ther surgery was certainly missing. Navigated 3D acquisition arrays. Certainly, more experience with real-time 3D imag-
was especially helpful for intraoperative planning of the ing will help as well. Finally, combination with a navigation
approach, whereas subsequent postoperative reconstructed system could overcome the persistent orientation problem.
3D resection control was not helpful in our experience. The surgeon could immediately see the US image in all three
After complete microscopic resection of the cavernoma, the familiar dimensions/planes without any time delay due to
surrounding tissue remained hyperechogenic in 3D US recon- acquisition or calculation. This would make neurosonogra-
structions and it thus imitated a significant residual lesion or phy a very strong competitive alternative in intraoperative
even the condition of no resection at all. We have used real- imaging.
time 2D US for resection control of cavernomas when using
navigated ultrasound. For resection control, real-time 3D US
is superior. The images are clearer and immediately available.
Conclusion
However, the image quality is not comparable to a high field
intraoperative MRI or high frequency multi-focus ultrasound
probes, and landmarking of the residual tumour for navigated These two presented intraoperative 3D US techniques offer
identification is still not possible. user-friendly and clinically useful imaging modalities and
represent further important steps in neurosonography. Ori-
entation in a one-platform solution with neuronavigation
remains superior to that in a real-time US system. However,
Standalone Use a larger clinical series should look deeper into these prelimi-
nary experiences.
One positive aspect of US technologies is their frequently Conflict of interest statement We declare that we have no conflict
mentioned low cost [9, 11, 12]. Neuronavigation has become of interest.
Acknowledgment The authors state that both US probes (IGSonic 9. Unsgaard G, Rygh GM, Selbekk T, Müller TB, Kolstad F, Lindseth F,
and real time 3D transducer) have been provided by the companies for Nagelhus Hernes TA (2006) Intra operative 3D ultrasound in neuro
research proposes. This is not the case for the navigation system surgery. Acta Neurochir 148:235 254
VectorVision or US system IU22, which were bought by the depart 10. Sure U, Benes L, Bozinov O, Woydt M, Tirakotai W, Bertalanffy H
ments. No further financial collaborations, consulting contracts or con (2005) Intraoperative landmarking of vascular anatomy by integ
flicts of interest exist. The reprint of the image in Fig. 3b has been ration of duplex and Doppler ultrasonography in image guided
approved by the company (Copyright Philips, Bothell, USA). surgery. Technical note. Surg Neurol 63:133 142
11. Tirakotai W, Miller D, Heinze S, Benes L, Bertalanffy H, Sure U
(2006) A novel platform for image guided ultrasound. Neurosur
gery 58:710 718
References 12. Unsgaard G, Ommedal S, Muller T, Gronningsaeter A,
Nagelhus Hernes TA (2002) Neuronavigation by intraoperative
three dimensional ultrasound: initial experience during brain
1. Bernays RL (2003) Intraoperative imaging in neurosurgery. MRI, CT, tumor resection. Neurosurgery 50:804 812
ultrasound. Introduction. Acta Neurochir Suppl 85:1 3 13. Sure U, Alberti O, Petermeyer M, Becker R, Bertalanffy H (2000)
2. Bonsanto MM, Staubert A, Wirtz CR, Tronnier V, Kunze S (2001) Advanced image guided skull base surgey. Surg Neurol 53:
Initial experience with an ultrasound integrated single rack neuro 563 572
navigation system. Acta Neurochir 143:1127 1132 14. Miller D, Heinze S, Tirakotai W, Bozinov O, Sürücü O, Benes L,
3. Cannon JW, Stoll JA, Sago IS, Knowles HB, Howe RD, Bertalanffy H, Sure U (2007) Is the image guidance of ultrasono
Dupont PE, Marx GR, del Nido PJ (2003) Real time three dimen graphy beneficial for neurosurgical routine? Surg Neurol 67:
sional ultrasound for guiding surgical tasks. Comput Aided Surg 579 587
8:82 90 15. Sure U, Gatscher S, Alberti O, Witte J, Bertalanffy H (2000)
4. Enchev Y, Bozinov O, Miller D, Tirakotai W, Heinze S, Benes L, Image guided duplex and Doppler ultrasound for microsurgery of
Bertalanffy H, Sure U (2006) Image guided ultrasonography for cerebral AVMs. Zentralbl Neurochir 61(Suppl 1):47 48 (abstract)
recurrent cystic gliomas. Acta Neurochir 148:1053 1063 16. Woydt M, Horowski A, Krauss J, Krone A, Soerensen N, Roosen K
5. Jödicke A, Deinsberger W, Erbe H, Kriete A, Böker D K (1998) (2002) Three dimensional intraoperative ultrasound of vascular
Intraoperative three dimensional ultrasonography: an approach to malformations and supratentorial tumors. J Neuroimaging 12:
register brain shift using multidimensional image processing. 28 34
Minim Invas Neurosurg 41:13 19 17. Unsgaard G, Ommedal S, Rygh OM, Lindseth F (2007) Operation
6. Jödicke A, Springer T, Böker D K (2004) Real time integration of of arteriovenous malformations assisted by stereoscopic naviga
ultrasound into neuronavigation: technical accuracy using a light tion controlled display of preoperative magnetic resonance angi
emitting diode based navigation system. Acta Neurochir ography and intraoperative ultrasound angiography. Neurosurgery
146:1211 1220 61(1 Suppl):407 415
7. Lindner D, Trantakis C, Renner C, Arnold S, Schmitgen A, 18. Gronningsaeter A, Kleven A, Ommedal S, Aarseth TE, Lie T,
Schneider J, Meixensberger J (2006) Application of intraoperative Lindseth F, Lango T, Unsgard G (2000) SonoWand, an
3D ultrasound during navigated tumor resection. Minim Invasive ultrasound based neuronavigation system. Neurosurgery 47:
Neurosurg 49:197 202 1373 1380
8. Lindseth F, Kaspersen JH, Ommedal S, Lango T, Bang J, Hokland J, 19. Winkler D, Lindner D, Strauss G, Richter A, Schober R,
Unsgaard G, Hernes TA (2003) Multimodal image fusion in ultra Meixensberger J (2006) Surgery of cavernous malformations
sound based neuronavigation: improving overview and interpreta with and without navigational support a comparative study.
tion by integrating preoperative MRI with intraoperative 3D Minim Invasive Neurosurg 49:15 19
ultrasound. Comput Aided Surg 8:49 69
Multimodality Integration
Integrated Intra-operative Room Design
Ivan Ng
Abstract The design of intraoperative suites require signi- indigo-cyanine green (ICG) angiography [9, 10] have made
ficant inputs from the neurosurgeons. Prior consideration their way into the operating rooms and considerable experi-
of specific surgical objectives before investment of capital ence with a wide array of systems have been accumulated.
resources will enable to surgeon to yield maximum value The move to incorporate various imaging modalities within
from the project. neurological surgery has been driven by the need to ensure
We describe the setup of the integrated neurosurgical that the desired operative endpoint or objective has been
centre at our institution which comprises of a hybrid high achieved. Examples would be the need to verify optimal
field MRI suite, an OR’s consisting of a multi-slice CT clip placement following aneurysm surgery, correct posi-
scanner and iso-C 3D respectively. The iCT and ioMRI tioning of screws in spinal instrumentation procedures and
OR’s carry ICG angiography capabilities. These ORs are more recently to ascertain brain tumour- normal brain inter-
linked to also the Novalis radiosurgery suites and outpatient face where the margins may not be readily apparent to the
clinics and offices to facilitate pre-surgical review, planning visible eye.
as well as treatment plans on a common interface via the The incorporation of these modalities in the operating
BRAINSUITE net. environment would in some instance require some degree
Design considerations include right sit-ing of imaging of configuration change of the operating suite as well as
equipment as well as a focus of ergonomics and design workflow modifications to accommodate this change from
features to maximize workflow. Whenever possible, stan- conventional neurosurgery. This may vary widely; it may
dard neurosurgical instrumentation is utilized. require little or no change in operating suite design and
With widespread availability of technology, neuro- construct; e.g. the use of 3D fluoroscopy to one which
imaging in the operating room may become more prevalent. requires significant planning, pre-design and construction
The surgeon is the lead individual in the team with regards to of an operating suite; e.g. the construction of a high field
planning and designing the ORs to accommodate the new intra-operative MRI facility.
imaging equipment. The construction of intra-operative operating rooms
(OR’s) in our institution for the purposes of neuro-
Keywords Intraoperative imaging Operating room design surgery was conceptualized to incorporate various forms of
neuro-imaging to cover and support the whole spectrum of
neurosurgical diseases into a pre-existing operating room
complex. In this manuscript, we describe the set-up of the
Introduction integrated neurosurgical centre focused on its various intra-
operative imaging options and objectives of design.
For more than a decade, advanced imaging modalities like
3D fluoroscopy [1], computed tomography (CT) [2, 3] and
magnetic resonance imaging( MRI) [4 6], catheter [7, 8] and Materials and Methods
I. Ng The integrated neurosurgical centre at the Singapore General

Departments of Neurosurgery at Singapore General Hospital and the
Hospital, Singapore comprised of three intra-operative OR’s
National Neuroscience Institute, Duke NUS Graduate Medical School,
Singapore and a Novalis radiosurgery suite (BrainLAB, Feldkirchen,
e mail: ivan.ng.h.b@sgh.com.sg Germany) which was built on a network platform that was
200 I. Ng
designed to enable seamless transfer of imaging information the scanning process without requiring any manual regis-
from workstations in the outpatient clinics and individual tration. For this process, a special MR compatible reference
neurosurgeons‘ office to the various OR. The design facili- structure and a reference star is attached to the upper half
tated easy access to the workstations of rendered software of the head coil and the head clamp. We routinely utilize
to allow for pre-surgical planning in any part of the hospital anatomical and functional as well as tractography images for
with wireless access and these data can be stored and navigation [4, 11 13]. The navigation workstation is placed
transferred to the navigational consoles within individual outside the radiofrequency shielded OR (Fig. 3). Three MRI
OR’s. These OR suites were specifically organized within compatible high resolution flat screen color monitors (57 in.)
the major operating theatre complex and accommodate the are available for the purposes of display of navigation guid-
building of a brand new operating room with a high field ance, live video streaming of the operating procedure and
magnetic resonance imaging machine (1.5 T; ioMRI); the neuro-images retrieved from the hospital in- house PACS
remaining two OR’s incorporate each an intra-operative system.
imaging modality, namely a 32 slice computed tomography
scanner on rails (iCT) and a 3 dimensional fluoroscopy
system( i-iso C 3D).respectively in two separate OR’s. Intra-Operative CT Scan (iCT) OR
The iCT OR was retrofitted into a pre-existing neurosurgical

Hybrid ioMRI Suite Setup OR and its dimensions are 8,350 mm7,000 mm3,000 mm
[approx. 59 m2] and incorporates a control room with the
dimensions 5,900 mm3,150 mm [approx. 19 m2]. The
The ioMRI suite measures 11,225 mm by 7,420 mm (ap-
main components of the OR includes a commercially avail-
proximately 83 m2 and was built on space previously used
able 32 multi-slice CT scanner (Somatom Sensation Open
within the major operating complex for post surgical recov-
Gliding Gantry; Siemens Medical Solutions, Forchheim),
ery. The control room which includes the server room was an
CT (32 slice; Siemens, Germany), a motorized radiolucent,
additional 7,865 mm3,500 mm. The layout of the
flexible rotating table (Trumpf, Pucheim, Germany), and a
operating suite is shown in Fig. 1. The high field 1.5 T
ceiling mounted frameless infrared-based neuronavigational
MRI(Magnetom Espree, Siemens; Erlangen, Germany) has
system (Vector Vision Sky; BrainLab) as in the ioMRI. The
a 70 cm wide bore open magnet with 30 cm of face space and
setup is represented in Figs. 3 and 4. The operating table
is 125 cm long and is housed within a pre-fabricated radio-
allows patient to be positioned in all positions except the
frequency-shielded cabin. The smaller footprint of the
sitting position and the CT scanner gantry moves over the
magnet allows for greater space conservation and the larger
patient during scanner by means of rails built on both sides of
bore allows for surgical procedures in the supine, prone or
the operating table (Fig. 3). As in the ioMRI suite, the scanner
lateral positions. The operating table (Trumpf, Pucheim,
is controlled by a workstation in a room next to the OR
Germany) is fully MRI compatible and in the default posi-
with direct visual contact and video surveillance. This allows
tion lies within the ‘‘sweet spot’’ of the operating room. It
for CT image acquisition without radiation exposure to
permits rotational movement of the patient enabling the head
personnel.
to be beyond the edge of the 5 Gauss safety line; additionally
the layout allows the provision to utilize neurosurgical pro-
cedures to be carried out with standard instrumentation.
When scanning is planned, the patient is placed in the mag- Brainsuite Network System
net by rotating the operating table. The patient’s head is held
in a MRI compatible ceramic head holder which allows 32 DICOM data generated from the MRI and CT scanner
coil elements to be seamlessly integrated into one examina- is networked between the BRAINSUITE net (BrainLAB,
tion and read out into eight independent receiver channels. A Feldkirchen, Germany) and the image guided surgery sys-
ceiling mounted microscope (Pentero, Zeiss, Oberkochen, tem which allows for automatic registration within the OR
Germany) with ICG angiography capabilities has a camera suite. The neuronavigation system also functions as a control
for tracking microscope movements is placed at a distance of terminal for BrainSUITE NET allowing for a common sim-
approximately 1.5 m in the 5-G perimeter (Fig. 2). We use ple interface for which to manipulate the data. We deployed
the Automatic Image Registration software & hardware a image data management system which allowed for video
(BrainLAB, Feldkirchen, Germany) which allows for auto- and still images like the surgical microscope and endoscope
matic registration of intra-operative images from the ioMRI as well as the OR light’s video camera and room cameras.
scanner. The intra-operative images are available for image This features a high end, firewall encapsulation of the
guided surgery at the navigation system immediately after complete subnetwork with controlled data link to the MR
Integrated Intra-operative Room Design 201
Fig. 1 Schematic overview of the IoMRI suite. The suite consists of the OR proper with its RF cabin; an anaesthetic preparation room, scrub nurse
preparation room, dedicated scrubbing area and the control and server room
scanner interface. Configurable connectivity to existing hos- Results

pital infrastructure like CT, MR. X-ray workstations and
PACS for DICOM image data transfer and even the possi- The casemix for the ioMRI and iCT suites from June 2008 to
bility of immediate remote service response was also catered July 2009 are shown in Table 1 and 2 respectively. Work-
for (Fig. 5). flow for the surgical team, nursing and anaesthetic team
202 I. Ng
RED ZONE YELLOW ZONE

1.5T MRI 50 GAUSS 5 GAUSS AND
STERILE DRAPE AND ABOVE ABOVE
SECURED TO
MR COMPATIBLE
DRIP STANDS
SECOND
SURGEON
FIRST
SURGEON
SCRUB
NURSE
ANAESTHETISTS
INSTRUMENT
TROLLEYS ANAESTHETIC
MACHINE
Fig. 2 General surgical layout within the ioMRI OR. Surgery is performed at the area which is beyond the 5G line (yellow perimeter). This allows
the use of conventional neurosurgical instrumentation. The arrangement allows for our usual OR configuration with regards to position of scrub
nurse and assistant
Fig. 3 Schematic overview of iCT and iso C3D suites

Fig. 4 iCT set up with a mobile

32 slice CT scanner on rails
with the radiolucent operating
table at its epicenter
Fig. 5 The BRAINSUITE net setup

204 I. Ng
Table 1 Casemix of ioMRI from May 2008 to June 2009 the use of the MRI has been shown to potentially benefit
ioMRI transphenoidal resection of pituitary tumours [14] intra-axial
Gliomas 37 tumours particularly gliomas [4, 11 13]; fluoroscopy with
Pituitary tumours 55 or without 3D reconstruction(as been demonstrated to be
AVMs 12 valuable with spinal instrumentation [10, 15, 16]. We there-
Metastasis 11
fore endeavored to build suites that will incorporate imaging
Biopsy 5
Total 120
modalities which may be best utilized to provide the informa-
tion necessary to the neurosurgeon for him to make critical
decisions and as an adjunct towards the smooth conduct of the
Table 2 Case of iCT from May 2008 to June 2009 operation. In this regard, we surmise that the MRI would be
iCT valuable for us in neuro-oncology [4 6, 12], the multi-slice
Spinal instrumentation 67 CT scanner would be useful in craniofacial procedures,
Cranial tumours 131 aneurysms with CT angiography [3], spinal instrumentation
Stereotaxic biopsy 33 and as a peri-operative surveillance tool; the iso C 3D
Aneurysms(icg & CTA) 7 fluoroscopy for spine [10, 15, 16] and ICG angiography
AVMs(iCT & CTA) 3 incorporated into the operating microscope in both the
Others (trauma, neuroendoscopy) 140 ioMRI and iCT so that vascular cases may be performed in
Total 381
either OR. With regard to the choice of the MRI, the three
choices would be a low field, 1.5 T and 3 T machines to
choose from [17 19]. We opted for a 1.5 T MRI for its
were essentially kept similar with minor modifications. With- superior image resolution and its ability to provide in addi-
in the OR, the surgical team and its spatial relationship with tion to structural anatomical images other sequences such
the scrub nurse and her instruments were unchanged to as tractography, diffusion weighted imaging, spectroscopy,
maintain efficiency and optimize coordination between the MR angiography and venography compared to the low field
two teams. Standard neurosurgical instruments were used systems. While there have been some 3 T systems employed
even in the ioMRI OR as the neurosurgical procedure was in the clinical setting, the advantage at this juncture over the
conducted just beyond the 5 G safety line. To increase 1.5 T machine for neuro-oncology has not been sufficiently
patient safety, the anaesthetic team deployment allows for investigated although there may be emerging evidence sug-
easy access to all vascular access and airway and physiolog- gest that the additional information with spectroscopy may
ical monitoring and can be continued remotely even during further aid glioma resections [17, 19, 20].
scanning with wireless devices or through cables running The design of these OR’s was also predicated on the fact
through the booms in the roof to the control rooms. All that it would be desirable not to modify too drastically the
medical equipment are placed on articulated booms to keep workflow or arrangements of the respective surgical and
heavy and bulky items out of harms way and away from the anaesthetic teams. We thus sought to ensure design would
imaging equipment. In the ioMRI OR, all equipment are suit our unique institutional requirements with regards to
secured in place. The OR floor is kept free of clutter as using the usual surgical instruments as well as relative posi-
much as possible. All the relevant OR;s utilize a line of tion of the surgeon and his assistant during the operations
sight neuronavigation system with a touchscreen with auto- well as scrub nurse positions to maintain and maximize
registration capabilities to minimize time wasted by trans- surgical efficiency. Optimization of ergonomics and work-
ferring new data into system and manually re-registering. flow also involved suspending whenever possible all surgi-
Duration where navigation is possible during the operation is cal equipment such as the electrocautery machine, suction,
maximized by having ceiling mounted line of sight naviga- craniotomy drills and endoscope trolleys onto booms sus-
tion systems. pended off the roof so as to minimize clutter on the floor.
The choice of the imaging system would largely be deter-
mined by the institutions’ casemix and in most instances, the
design and conceptualization would best be served by hav-
Discussion ing surgeons with a vested interest in utilizing the techno-
logy leading the project design and implementation. This
Various options for intra-operative modalities have got sig- would ensure that the ORs are built to specifications which
nificant different properties and capabilities to visualize would best facilitate and improve on contemporary practice.
neuroanatomical and potentially neurophysiological entities While turnkey solutions are available commercially, indivi-
and the current state of technology has demonstrated value dualized planning is still valuable to maximize efficiency
in individual categories of neurosurgical procedures; e.g., and ensure that the new ORs are appropriately used.
Conclusions 8. Tang G, Cawley CM, Dion JE, Barrow DL (2002) Intraoperative

angiography during aneurysm surgery: a prospective evaluation of
efficacy. J Neurosurg 96:993 999
The design and building of hybrid intraoperative OR’s 9. Dashti R, Laakso A, Niemela M, Porras M, Hernesniemi J (2009)
despite the increasing availability of turnkey solutions will Microscope integrated near infrared indocyanine green video
still need to be led by neurosurgeons with careful thought put angiography during surgery of intracranial aneurysms: the Helsinki
towards designing the OR’s to suit individual needs within experience. Surg Neurol 71:543 550, discussion 550
10. Hott JS, Papadopoulos SM, Theodore N, Dickman CA, Sonntag
the institution. VK (2004) Intraoperative Iso C C arm navigation in cervical spinal
surgery: review of the first 52 cases. Spine (Phila Pa 1976)
Conflict of interest statement We declare that we have no conflict of 29:2856 2860
interest. 11. Nimsky C, Ganslandt O, Fahlbusch R (2007) Implementation
of fiber tract navigation. Neurosurgery 61:306 317, discussion
317 318
12. Nimsky C, Ganslandt O, Hastreiter P, Wang R, Benner T,
References Sorensen AG, Fahlbusch R (2007) Preoperative and intraoperative
diffusion tensor imaging based fiber tracking in glioma surgery.
Neurosurgery 61:178 185, discussion 186
1. Stubig T, Kendoff D, Citak M, Geerling J, Khalafi A, Krettek C, 13. Nimsky C, von Keller B, Schlaffer S, Kuhnt D, Weigel D,
Hufner T (2009) Comparative study of different intraoperative 3 D Ganslandt O, Buchfelder M (2009) Updating navigation with
image intensifiers in orthopedic trauma care. J Trauma 66:821 830 intraoperative image data. Top Magn Reson Imaging 19:197 204
2. Engle DJ, Lunsford LD (1987) Brain tumor resection guided by 14. Nimsky C, von Keller B, Ganslandt O, Fahlbusch R (2006) Intrao
intraoperative computed tomography. J Neurooncol 4:361 370 perative high field magnetic resonance imaging in transsphenoidal
3. Uhl E, Zausinger S, Morhard D, Heigl T, Scheder B, Rachinger W, surgery of hormonally inactive pituitary macroadenomas. Neur
Schichor C, Tonn JC (2009) Intraoperative computed tomography osurgery 59:105 114, discussion 105 114
with integrated navigation system in a multidisciplinary operating 15. Holly LT, Foley KT (2003) Intraoperative spinal navigation. Spine
suite. Neurosurgery 64:231 239, discussion 239 240 (Phila Pa 1976) 28:S54 S61
4. Nimsky C, Ganslandt O, Kober H, Buchfelder M, Fahlbusch R 16. Holly LT, Foley KT (2007) Image guidance in spine surgery.
(2001) Intraoperative magnetic resonance imaging combined with Orthop Clin North Am 38:451 461, abstract viii
neuronavigation: a new concept. Neurosurgery 48:1082 1089, dis 17. Kim PD, Truwit CL, Hall WA (2009) Three tesla high field appli
cussion 1089 1091 cations. Neurosurg Clin N Am 20:173 178
5. Sutherland GR, Kaibara T, Louw D, Hoult DI, Tomanek B, 18. Lewin JS, Metzger AK (2001) Intraoperative MR systems.
Saunders J (1999) A mobile high field magnetic resonance system Low field approaches. Neuroimaging Clin N Am 11:611 628
for neurosurgery. J Neurosurg 91:804 813 19. Pamir MN, Peker S, Ozek MM, Dincer A (2006) Intraoperative
6. Truwit CL, Hall WA (2006) Intraoperative magnetic resonance MR imaging: preliminary results with 3 tesla MR system. Acta
imaging guided neurosurgery at 3 T. Neurosurgery 58:ONS 338 Neurochir Suppl 98:97 100
ONS 345, discussion ONS 345 ONS 346 20. Pamir MN, Ozduman K, Dincer A, Yildiz E, Peker S, Ozek MM
7. Kivisaari RP, Porras M, Ohman J, Siironen J, Ishii K, Hernesniemi J (2010) First intraoperative, shared resource, ultrahigh field 3 Tesla
(2004) Routine cerebral angiography after surgery for saccular magnetic resonance imaging system and its application in low grade
aneurysms: is it worth it? Neurosurgery 55:1015 1024 glioma resection. J Neurosurg 112(1):57 69
Multimodal Navigation Integrated with Imaging
Christopher Nimsky, Daniela Kuhnt, Oliver Ganslandt, and Michael Buchfelder
Abstract Intraoperative high-field MRI in combination estimation by the surgeon, intraoperative imaging allows
and close integration with microscope-based navigation an objective evaluation of the intraoperative situation, thus
serving as a common interface for the presentation of multi- acting as quality control during surgery [1 6]. In addition
modal data in the surgical field seems to be one of the to intraoperative imaging, an integral part of our concept
most promising surgical setups allowing avoiding unwanted of computer aided surgery is the simultaneous application
tumor remnants while preserving neurological function. of navigation [5, 7, 8]. This integrated navigation allows
Multimodal navigation integrates standard anatomical, struc- essentially visualizing the results of pre- and intraoperative
tural, functional, and metabolic data. Navigation achieves imaging in the surgical field, so that the image data provide
visualizing the initial extent of a lesion with the concomitant an immediate feedback. The main goal is to prevent
identification of neighboring eloquent brain structures, as well increased neurological deficits despite increased resections
as, providing a tool for a direct correlation of histology and that might result from the attempt to remove initially over-
multimodal data. With the help of intraoperative imaging looked tumor remnants that are detected by intraoperative
navigation data can be updated, so that brain shift can be imaging.
compensated for and initially missed tumor remnants can be In standard navigation, also known as frameless stereo-
localized reliably. taxy, the real space of the surgical field is registered to the
3-D image space, which is based on anatomical data only.
Keywords Brain shift Intraoperative MRI MRI Neuro- We prefer the application of microscope-based navigation,
navigation where the extent and localization of a tumour is superim-
posed on the surgical field through contours using the heads-
up display technology of the modern operating microscopes.
Standard navigation, based on anatomical information only,
Introduction which has become a routine tool in many neurosurgical
departments, was enhanced by the integration of further
The possibility to objectively determine the extent of tu- information from other modalities resulting meanwhile in
mour removal during surgery is highly advantageous. Thus, the so-called multimodal navigation.
intraoperative imaging has gained increasing interest in the Functional navigation, integrating preoperative data
last decade. If a resection is incomplete, one can attempt to from magnetoencephalography (MEG) [9 11] and function-
remove the tumour residues that were initially missed al magnetic resonance imaging (fMRI) [5, 12] to define
during the same operation. In contrast to a subjective the localization of cortical eloquent brain areas, such as the
motor and speech areas, was the first step in the direction
of modern multimodal navigation. Functional navigation
allowed more thorough resections of tumours in risk zones
C. Nimsky (*) and D. Kuhnt with low morbidity. Integration of diffusion tensor imaging
Department of Neurosurgery, University Marburg, Baldingerstrasse (DTI) data delineating the course of major white matter tracts
35033 Marburg, Germany
extended this concept also to subcortical areas [13 16],
e mail: nimsky@med.uni marburg.de
while the co-registration of PET data and information
O. Ganslandt and M. Buchfelder
Department of Neurosurgery, University Erlangen Nuremberg, Erlangen, from MR spectroscopy (MRS) added metabolic information
Germany leading to true multimodal navigation [17 22].
208 C. Nimsky et al.
Navigation in an Environment Integration of Multimodal Data

with Intraoperative Imaging
It is absolutely mandatory to combine the goal of maximum
There are several concepts implementing the combination resection with the goal of preservation of function. Intra-
of navigation and intraoperative imaging. One possibility is operative imaging helps on one side to maximize the extent
that the MR scanner serves as a navigational device per se, as of resection, but on the other side it also allows in combina-
it was the basic principle in the 0.5 T double-doughnut GE tion with functional multimodal navigation a minimization
scanner concept, where the patient was operated directly in of postoperative neurological deficits [13, 29]. With the
the scanner [1, 23, 24]. Since surgical space and imaging advances in surgical techniques and perioperative technolo-
space are identical, an instrument in the surgical space can gy, it is now possible to maximally resect malignant intrinsic
be tracked in the image space without much additional glial neoplasms, even close to functionally critical areas,
effort. Direct navigation in the MR scanner is often based with minimally increased morbidity. Recent studies have
on realtime imaging, like in the so-called prospective stereo- demonstrated a survival advantage of these lesions with a
taxy, a method for trajectory alignment for placements of resection extent of 98% or greater, particularly in younger
catheters or sampling biopsies [25 27]. patients with good Karnofsky scores [30].
Alternative attempts to integrate intraoperative MRI and Functional navigation, i.e. integrating functional data into
navigation result in a standard navigation setup because anatomical navigational datasets, is an important add-on to
image space and surgical space are not identical, necessitat- intraoperative MRI since it prevents too extensive resec-
ing some kind of patient registration. Most setups implement tions, which would otherwise result in new neurological
navigation at the 5 G line, so that standard non-MR-compatible deficits. Meanwhile, data from MEG and fMRI can be rou-
instruments can be used. Ceiling mounted solutions of the tinely integrated in functional navigation allowing identifi-
navigation camera and screens [5] are an optimal solution in cation of eloquent brain areas such as the motor area and
the intraoperative scenario, since placing a standard naviga- speech related areas [10, 12].
tion system just close to the MR scanner might increase the In a retrospective study we analyzed how the decision
risk of potential magnetic accidents. High navigation accu- for glioma resection was influenced by MEG [9]. In a time
racy is a prerequisite if the navigation information is to be period of 5 consecutive years we had investigated every
used at critical steps during the resection of a tumor. Among patient proposed for surgery, harbouring a lesion adjacent
all errors contributing to the overall navigation accuracy, the to eloquent brain areas. Altogether 191 patients were exam-
initial patient registration process is mostly prone to errors. ined, 119 of them with supratentorial gliomas. About every
The most common strategy for patient registration relies on forth patient (26.8%) yielded a severe possible danger of
placement of skin-adhesive fiducials, which can be detected postoperative neurological morbidity according to MEG and
in the images, so that their position in virtual and real physi- thus was not considered being a good candidate for surgery.
cal space can be correlated to define the registration coor- This corresponds well to other published data where 12 out
dinate system. Automatic registration setups, allowing an of 40 investigated patients (30%) with tumours and vascular
user-independent registration of patient space and image malformations underwent non-surgical therapy according to
space, try to reduce the user dependent errors [28]. the MEG findings [31]. When functional data were used
Our concept of automatic registration is based on a regis- in combination with frameless stereotactic devices the post-
tration matrix containing markers at predefined positions operative morbidity was as low as 2.3%. Overall morbidity
which are visible during the imaging process. Combining however was 6.8%. These data reflect the beneficial effects
the information where the registration matrix is in relation to of functional navigation in comparison to data of other
the reference array and how the detected markers in the studies with morbidity rates varying from 6 to 31.7%
images relate to the actual markers of the registration matrix [32 36]. These data can also be interpreted as a result of a
allows calculating a transformation matrix, so that then the more careful patient selection through the help of advanced
relation between image space and physical/surgical space is preoperative brain mapping. Preoperative identification of
defined and navigation can be used. An additional skin fidu- eloquent brain areas has an impact in the risk evaluation
cial that is not used for the registration process is localized in glioma surgery, as well as functional navigation reduces
after patient registration to document a target registration the risk for postoperative neurological deficits. Besides
error, which is typically in the range between 0.3 and identification of the motor strip, the localisation of language
2.5 mm. Phantom studies resulted in median localization areas is of great clinical impact [37, 38].
errors between 0.88 and 2.13 mm for the automatic registra- Functional data from MEG and fMRI only localize func-
tion approach, which was at least not worse, in most test series tional areas at the brain surface. However, neurological
even significantly better, than that of the standard registration deficits can also be caused during tumor resection by dam-
no matter whether 4 or 7 fiducial markers were used [28]. aging of deeper structures, such as major white matter tracts.
Multimodal Navigation Integrated with Imaging 209
Diffusion tensor imaging (DTI) can be used not only to data only the accuracy decreases with the course of surgery.
delineate tumor borders, but also to display the course of Intraoperative imaging offers a possibility to compensate for
major white matter tracts, such as the pyramidal tract. The the effects of brain shift, because it provides a virtual repro-
knowledge of the course of these tracts in relationship to a duction of the actual intraoperative physical reality, both in
tumor helps to reduce or even prevent new postoperative regard of brain deformation and on the actual extent of tumor
neurological deficits [39, 40]. The registration of diffusion removal [24, 47 49]. Updating of the navigation by intra-
data with the navigational data [16, 41, 42] facilitates operative registration of the intraoperative image data had
the intraoperative preservation of these eloquent structures. been a cumbersome process in the initial low-field MRI
A prerequisite is that intraoperative changes of the brain systems. Bone fiducial markers, which had to be placed
anatomy, known as brain shift, are taken into account. In around the craniotomy, were hardly detectable in the intra-
contrast to the use of fMRI and MEG, brain shift clinically operative images, as well as the whole update procedure was
impacts the DTI data to a much more relevant extent, because a quite time-consuming process [48, 49]. This mainly
the intraoperative shifting of cortical areas during surgery can explains why only in 16 out of 330 patients investigated
be well detected by direct observation with the operating with low-field MRI an actual navigation update was per-
microscope, however changes in the depth of a resection formed, despite intraoperative imaging had detected tumor
cavity, close to major white matter tracts, is nearly undetect- remnants e.g. in the gliomas that could undergo further
able for the neurosurgeon during tumor resection. Intraopera- resection in about 26% [48, 50].
tive DTI is a convenient possibility to visualize this shifting Integrating high-field MRI and microscope-based navi-
during tumor resection [15, 43]. As a consequence of brain gation allowed facilitating this intraoperative update proce-
shift preoperative functional data are no longer valid during dure clearly [5]. One possibility for intraoperative image
the course of tumor removal, so navigation can no longer registration is to apply the automatic registration matrix by
relied on, if this shifting is not compensated for. Therefore, attaching it to the upper part of the head coil, like it is done
it is necessary that not only intraoperative anatomical data are for the initial patient registration process with preoperative
used to compensate for the effects of brain shift but also image data [28]. Alternatively navigation updating is also
functional data have to be updated. Intraoperative acquisition possible without repeated patient registration. This approach
of DTI data enables intraoperative fiber tracking to visualize is based on a rigid registration of the intraoperative image
how a tumor remnant is localized in relation to major white data with the preoperative image data, subsequent segmen-
matter tracts [15, 43]. Even intraoperative fMRI applying tation of the tumor remnant, and final restoring of the initial
electrical stimulation of median and tibial nerves as a passive patient registration. Thus, the registration coordinate system
stimulation paradigm is possible and enables identification of of the preoperative image data is applied on the intraopera-
the somatosensory cortex [44]. tive images, serving as an immediate intraoperative image
Besides functional and structural data further information update. Updated image data allow a reliable identification
is available for a multimodal navigation setup. PET, MRS, of a tumor remnant or correction of a catheter position.
and diffusion weighted imaging may provide information on Microscope-based image injection with the direct visualiza-
the diffuse tumor border. Integration of metabolic maps into tion of the segmented tumor remnant in the surgical field
the neuronavigation datasets enables a spatial correlation of plays a crucial role in the precise localization and orienta-
metabolic data and histopathological findings [19, 45]. tion in the resection cavity. Histological analysis of the
Whether these techniques can also be used intraoperatively, extended resections proved pathological tissue in all cases;
so that these data can also be updated, is under investiga- there were no false positive findings. This is also mainly
tion. Furthermore, upcoming techniques such as MR-based due to the fact that only areas of reliably identified tumor
molecular imaging may find its role in the intraoperative remnants were segmented and used for updating. The side-
imaging armamentarium. by-side analysis of pre- and intraoperative images greatly
facilitates image interpretation and excludes misinterpreta-
tions due to surgically induced changes at the resection
border. Repeated landmark checks have proved that the
rigid registration approach is a reliable approach to update
Navigation Updating by Intraoperative
the navigation system without a repeated patient registra-
Image Data tion. The rigid registration algorithm is robust enough to
accomplish a registration that is not sensitive to the effects
Tumor removal, brain swelling, the use of brain retractors, of brain shift and the actual reduction of the tumor mass.
and cerebrospinal-fluid drainage all result in intraoperative This could be shown by analyzing the registration accuracy
brain deformation, which is known as brain shift [24, 46]. of structures that are not affected by brain shift. An exten-
Thus, in navigation systems relying on preoperative image sive analysis comparing the automatic registration, which is
used to register pre- and intraoperative images, with an Nevertheless, to prevent a mis-registration a visual control
independent reference registration showed, that the registra- after rigid registration of pre- and intraoperative images is
tion error is below 2 mm even in a worst case scenario [51]. mandatory. Figures 1 3 depict an illustrative example of
Fig. 1 22 years old male patient

with a left sided parietal pilocytic
astrocytoma, (a) preoperative
image depicting the large cystic
component, (b) intraoperative
imaging after completion of
tumor resection depicting
complete tumor removal (both:
contrast enhanced T1 weighted
images)
Fig. 2 Same patient as in Fig. 1; an update with intraoperative images was performed twice a/b depict the first intraoperative update with a
co registered display of pre and first intraoperative (iop1) images, while the tumor remnant was segmented in the iop1 images (b) and also
displayed in the preoperative images (a) visualizing the distinct shift; c/d depict the second update after further tumor resection and repeated
imaging, co registered are first and second intraoperative imaging (iop2), the tumor remnant was segmented in the iop2 images (d) and also
visualized in the iop1 images (c) (all images T2 weighted axial scans displayed with the navigation software, white arrows depict the segmented
tumor remnants)
Fig. 3 Same patient as in Figs. 1 and 2 depicting the course of resection in two slice positions that were identical in all 4 imaging sessions (pre: a/f,
iop1: b/f, iop2: c/g, iop3 d/h, all images axial T2 weighted, slice thickness 3 mm, difference between slice positions row a d to row e h: 9 mm),
note the different portions of brain shift first after opening the large tumor cysts (difference a/e to b/f) and additionally after first updating
collapsing of the left frontal ventricular horn (f to g)
repeated intraoperative updating during resection of a pilo- 3-D ultrasound [54 56]. Whether the image quality to evalu-
cytic astrocytoma. ate the extent of a glioma resection is really equivalent among
Updating the navigation system with intraoperative high- the different imaging modalities is still discussed controver-
field MR image data at the moment seems to be the most sially. Mathematical models describing the deformation of the
reliable method to compensate for the effects of brain shift. brain during surgery might help in the process of deforming
In contrast to previous setups this framework is also open to preoperative image data [57 61]. However, it is important
update multimodal information. Functional data, such as that some sparse data describing the actual intraoperative 3-D
fMRI or DTI data can also be acquired intraoperatively situation serve as input for the mathematical models, so that
and directly used for intraoperative updating, which in the they are able to adjust high-quality preoperative data to repre-
clinical routine might be a time-consuming effort, especially sent the intraoperative reality [59, 61 63]. If such a workflow
when in case of e.g. visualization of speech connecting fiber is established intraoperative high-field MRI with anatomical
tracts some sophisticated time-consuming non-standard and functional imaging possibilities would be the ideal tool to
tracking algorithms have to be applied. validate and refine these models.
Alternatively non-linear registration techniques or sophis-
ticated techniques from pattern recognition analysis may
allow a matching of preoperative MR data sets containing
functional information with intraoperative MR image Conclusion
volumes [52, 53]. This might also be an approach in cases
where intraoperative MRI is not available, but other imaging Intraoperative high-field MRI in combination and close inte-
modalities provide intraoperative 3-D information about gration with microscope-based navigation serving as a com-
the brain configuration, so that high-resolution multi-modality mon interface for the presentation of multimodal data in the
data can be registered non-linearly onto the ‘‘low-quality’’ surgical field seems to be one of the most promising surgical
intraoperative data. Such an alternative to intraoperative MRI setups allowing avoiding unwanted tumor remnants while
might be intraoperative ultrasound, especially intraoperative preserving neurological function. Multimodal navigation
integrates standard anatomical, structural, functional, and 13. Nimsky C, Ganslandt O, Fahlbusch R (2005) 1.5 T: intraoperative
metabolic data. Navigation achieves visualizing the initial imaging beyond standard anatomic imaging. Neurosurg Clin N Am
16:185 200, vii
extent of a lesion with the concomitant identification of 14. Nimsky C, Ganslandt O, Fahlbusch R (2006) Implementation of
neighboring eloquent brain structures, as well as, providing fiber tract navigation. Neurosurgery 58:292 304
a tool for a direct correlation of histology and multimodal 15. Nimsky C, Ganslandt O, Hastreiter P, Wang R, Benner T, Sorensen
data. With the help of intraoperative imaging navigation AG, Fahlbusch R (2005) Preoperative and intraoperative diffusion
tensor imaging based fiber tracking in glioma surgery. Neurosur
data can be updated, so that brain shift can be compensated gery 56:130 138
for and initially missed tumor remnants can be localized 16. Nimsky C, Ganslandt O, Merhof D, Sorensen AG, Fahlbusch R
reliably. (2006) Intraoperative visualization of the pyramidal tract by
diffusion tensor imaging based fiber tracking. Neuroimage 30:
Conflict of interest statement Ch. Nimsky is scientific consultant for 1219 1229
Brainlab. 17. Ganslandt O, Stadlbauer A, Fahlbusch R, Kamada K, Buslei R,
Blumcke I, Moser E, Nimsky C (2005) Proton magnetic resonance
spectroscopic imaging integrated into image guided surgery: cor
relation to standard magnetic resonance imaging and tumor cell
density. Neurosurgery 56:291 298
References 18. Stadlbauer A, Ganslandt O, Buslei R, Hammen T, Gruber S, Moser
E, Buchfelder M, Salomonowitz E, Nimsky C (2006) Gliomas:
histopathologic evaluation of changes in directionality and magni
1. Black PM, Moriarty T, Alexander E III, Stieg P, Woodard EJ, tude of water diffusion at diffusion tensor MR imaging. Radiology
Gleason PL, Martin CH, Kikinis R, Schwartz RB, Jolesz FA 240:803 810
(1997) Development and implementation of intraoperative mag 19. Stadlbauer A, Moser E, Gruber S, Nimsky C, Fahlbusch R,
netic resonance imaging and its neurosurgical applications. Neuro Ganslandt O (2004) Integration of biochemical images of a tumor
surgery 41:831 845 into frameless stereotaxy achieved using a magnetic resonance
2. Hall WA, Kowalik K, Liu H, Truwit CL, Kucharezyk J (2003) imaging/magnetic resonance spectroscopy hybrid data set. J Neu
Costs and benefits of intraoperative MR guided brain tumor resec rosurg 101:287 294
tion. Acta Neurochir Suppl 85:137 142 20. Stadlbauer A, Nimsky C, Buslei R, Pinker K, Gruber S, Hammen
3. Hall WA, Liu H, Martin AJ, Pozza CH, Maxwell RE, Truwit CL T, Buchfelder M, Ganslandt O (2007) Proton magnetic resonance
(2000) Safety, efficacy, and functionality of high field strength spectroscopic imaging in the border zone of gliomas: correlation of
interventional magnetic resonance imaging for neurosurgery. metabolic and histological changes at low tumor infiltration initial
Neurosurgery 46:632 642 results. Invest Radiol 42:218 223
4. Nimsky C, Ganslandt O, Fahlbusch R (2005) Comparing 0.2 tesla 21. Stadlbauer A, Nimsky C, Buslei R, Salomonowitz E, Hammen T,
with 1.5 tesla intraoperative magnetic resonance imaging analysis Buchfelder M, Moser E, Ernst Stecken A, Ganslandt O (2007)
of setup, workflow, and efficiency. Acad Radiol 12:1065 1079 Diffusion tensor imaging and optimized fiber tracking in glioma
5. Nimsky C, Ganslandt O, Von Keller B, Romstock J, Fahlbusch R patients: histopathologic evaluation of tumor invaded white matter
(2004) Intraoperative high field strength MR imaging: implemen structures. Neuroimage 34:949 956
tation and experience in 200 patients. Radiology 233:67 78 22. Stadlbauer A, Prante O, Nimsky C, Salomonowitz E, Buchfelder M,
6. Sutherland GR, Kaibara T, Louw D, Hoult DI, Tomanek B, Kuwert T, Linke R, Ganslandt O (2008) Metabolic imaging of
Saunders J (1999) A mobile high field magnetic resonance system cerebral gliomas: spatial correlation of changes in O (2 18F fluor
for neurosurgery. J Neurosurg 91:804 813 oethyl) L tyrosine PET and proton magnetic resonance spectro
7. Nimsky C, Ganslandt O, Kober H, Buchfelder M, Fahlbusch R scopic imaging. J Nucl Med 49:721 729
(2001) Intraoperative magnetic resonance imaging combined with 23. Black PM, Alexander E III, Martin C, Moriarty T, Nabavi A,
neuronavigation: a new concept. Neurosurgery 48:1082 1091 Wong TZ, Schwartz RB, Jolesz F (1999) Craniotomy for tumor
8. Steinmeier R, Fahlbusch R, Ganslandt O, Nimsky C, Buchfelder M, treatment in an intraoperative magnetic resonance imaging unit.
Kaus M, Heigl T, Lenz G, Kuth R, Huk W (1998) Intraoperative Neurosurgery 45:423 433
magnetic resonance imaging with the magnetom open scanner: 24. Nabavi A, Black PM, Gering DT, Westin CF, Mehta V,
concepts, neurosurgical indications, and procedures: a preliminary Pergolizzi RS Jr, Ferrant M, Warfield SK, Hata N, Schwartz RB,
report. Neurosurgery 43:739 748 Wells WM 3rd, Kikinis R, Jolesz FA (2001) Serial intraopera
9. Ganslandt O, Buchfelder M, Hastreiter P, Grummich P, tive magnetic resonance imaging of brain shift. Neurosurgery
Fahlbusch R, Nimsky C (2004) Magnetic source imaging supports 48:787 798
clinical decision making in glioma patients. Clin Neurol Neurosurg 25. Hall WA, Liu H, Truwit CL (2000) Navigus trajectory guide.
107:20 26 Neurosurgery 46:502 504
10. Ganslandt O, Fahlbusch R, Nimsky C, Kober H, Möller M, Stein 26. Truwit CL, Hall WA (2006) Intraoperative magnetic resonance
meier R, Romstöck J, Vieth J (1999) Functional neuronavigation imaging guided neurosurgery at 3 T. Neurosurgery 58:ONS338
with magnetoencephalography: outcome in 50 patients with ONS345, discussion ONS345 ONS346
lesions around the motor cortex. J Neurosurg 91:73 79 27. Truwit CL, Liu H (2001) Prospective stereotaxy: a novel method of
11. Ganslandt O, Steinmeier R, Kober H, Vieth J, Kassubek J, trajectory alignment using real time image guidance. J Magn
Romstock J, Strauss C, Fahlbusch R (1997) Magnetic source imag Reson Imaging 13:452 457
ing combined with image guided frameless stereotaxy: a new meth 28. Rachinger J, von Keller B, Ganslandt O, Fahlbusch R, Nimsky C
od in surgery around the motor strip. Neurosurgery 41:621 628 (2006) Application accuracy of automatic registration in frameless
12. Nimsky C, Ganslandt O, Kober H, Moller M, Ulmer S, Tomandl B, stereotaxy. Stereotact Funct Neurosurg 84:109 117
Fahlbusch R (1999) Integration of functional magnetic resonance 29. Nimsky C, Ganslandt O, Fahlbusch R (2004) Functional neurona
imaging supported by magnetoencephalography in functional neu vigation and intraoperative MRI. Adv Tech Stand Neurosurg
ronavigation. Neurosurgery 44:1249 1256 29:229 263
30. Hentschel SJ, Sawaya R (2003) Optimizing outcomes with maxi 47. Hastreiter P, Rezk Salama C, Soza G, Bauer M, Greiner G,
mal surgical resection of malignant gliomas. Cancer Control Fahlbusch R, Ganslandt O, Nimsky C (2004) Strategies for brain
10:109 114 shift evaluation. Med Image Anal 8:447 464
31. Hund M, Rezai AR, Kronberg E, Cappell J, Zonenshayn M, 48. Nimsky C, Ganslandt O, Hastreiter P, Fahlbusch R (2001) Intra
Ribary U, Kelly PJ, Llinas R (1997) Magnetoencephalographic operative compensation for brain shift. Surg Neurol 56:357 365
mapping: basic of a new functional risk profile in the selection of 49. Wirtz CR, Bonsanto MM, Knauth M, Tronnier VM, Albert FK,
patients with cortical brain lesions. Neurosurgery 40:936 943 Staubert A, Kunze S (1997) Intraoperative magnetic resonance
32. Ammirati M, Galicich JH, Arbit E, Liao Y (1987) Reoperation imaging to update interactive navigation in neurosurgery: method
in the treatment of recurrent intracranial malignant gliomas. and preliminary experience. Comput Aided Surg 2:172 179
Neurosurgery 21:607 614 50. Nimsky C, Ganslandt O, Tomandl B, Buchfelder M, Fahlbusch R
33. Black PM (1999) Surgery for cerebral gliomas: past, present and (2002) Low field magnetic resonance imaging for intraopera
future. In: Howard MA III, Elliott JP, Haglund MM, McKhann GM tive use in neurosurgery: a 5 year experience. Eur Radiol 12:
II (eds) Clinical neurosurgery, vol 47. Lippincott Williams & 2690 2703
Wilkins, Boston, pp 21 45 51. Veyrat A (2005) Automatic fusion of pre and intraoperative
34. Cabantog AM, Bernstein M (1994) Complications of first craniot patienta data. A statistical evaluation of accuracy. Diploma thesis,
omy for intra axial brain tumour. Can J Neurol Sci 21:213 218 Technical University Munich
35. Ciric I, Ammirati M, Vick N, Mikhael M (1987) Supratentorial 52. Archip N, Clatz O, Whalen S, Kacher D, Fedorov A, Kot A,
gliomas: surgical considerations and immediate postoperative Chrisochoides N, Jolesz F, Golby A, Black PM, Warfield SK
results. Gross total resection versus partial resection. Neurosurgery (2007) Non rigid alignment of pre operative MRI, fMRI, and
21:21 26 DT MRI with intra operative MRI for enhanced visualization
36. Fadul C, Wood J, Thaler H, Galicich J, Patterson RH Jr, Posner JB and navigation in image guided neurosurgery. Neuroimage
(1988) Morbidity and mortality of craniotomy for excision of 35:609 624
supratentorial gliomas. Neurology 38:1374 1379 53. Wolf M, Vogel T, Weierich P, Niemann H, Nimsky C (2001)
37. Grummich P, Nimsky C, Pauli E, Buchfelder M, Ganslandt O Automatic transfer of preoperative fMRI markers into intraopera
(2006) Combining fMRI and MEG increases the reliability of tive MR images for updating functional neuronavigation. IEICE
presurgical language localization: a clinical study on the difference Trans Inf Syst E84 D:1698 1704
between and congruence of both modalities. Neuroimage 32: 54. Comeau RM, Sadikot AF, Fenster A, Peters TM (2000) Intraopera
1793 1803 tive ultrasound for guidance and tissue shift correction in image
38. Kober H, Moller M, Nimsky C, Vieth J, Fahlbusch R, Ganslandt O guided neurosurgery. Med Phys 27:787 800
(2001) New approach to localize speech relevant brain areas and 55. Letteboer MM, Willems PW, Viergever MA, Niessen WJ (2005)
hemispheric dominance using spatially filtered magnetoencepha Brain shift estimation in image guided neurosurgery using 3 D
lography. Hum Brain Mapp 14:236 250 ultrasound. IEEE Trans Biomed Eng 52:268 276
39. Clark CA, Barrick TR, Murphy MM, Bell BA (2003) White matter 56. Tirakotai W, Miller D, Heinze S, Benes L, Bertalanffy H, Sure U
fiber tracking in patients with space occupying lesions of the (2006) A novel platform for image guided ultrasound. Neurosur
brain: a new technique for neurosurgical planning? Neuroimage gery 58:710 718
20:1601 1608 57. Arbel T, Morandi X, Comeau RM, Collins DL (2004) Automatic
40. Hendler T, Pianka P, Sigal M, Kafri M, Ben Bashat D, Constantini S, non linear MRI ultrasound registration for the correction of intra
Graif M, Fried I, Assaf Y (2003) Delineating gray and white matter operative brain deformations. Comput Aided Surg 9:123 136
involvement in brain lesions: three dimensional alignment of 58. Coenen VA, Krings T, Weidemann J, Hans FJ, Reinacher P,
functional magnetic resonance and diffusion tensor imaging. Gilsbach JM, Rohde V (2005) Sequential visualization of brain
J Neurosurg 99:1018 1027 and fiber tract deformation during intracranial surgery with three
41. Coenen VA, Krings T, Mayfrank L, Polin RS, Reinges MH, Thron dimensional ultrasound: an approach to evaluate the effect of brain
A, Gilsbach JM (2001) Three dimensional visualization of the shift. Neurosurgery 56:133 141
pyramidal tract in a neuronavigation system during brain tumor 59. Lunn KE, Paulsen KD, Lynch DR, Roberts DW, Kennedy FE,
surgery: first experiences and technical note. Neurosurgery Hartov A (2005) Assimilating intraoperative data with brain shift
49:86 93 modeling using the adjoint equations. Med Image Anal 9:281 293
42. Nimsky C, Grummich P, Sorensen AG, Fahlbusch R, Ganslandt O 60. Rasmussen IA Jr, Lindseth F, Rygh OM, Berntsen EM, Selbekk T,
(2005) Visualization of the pyramidal tract in glioma surgery by Xu J, Nagelhus Hernes TA, Harg E, Haberg A, Unsgaard G (2007)
integrating diffusion tensor imaging in functional neuronavigation. Functional neuronavigation combined with intra operative 3D
Zentralbl Neurochir 66:133 141 ultrasound: initial experiences during surgical resections close to
43. Nimsky C, Ganslandt O, Hastreiter P, Wang R, Benner T, Sorensen A, eloquent brain areas and future directions in automatic brain
Fahlbusch R (2005) Intraoperative diffusion tensor MR imaging: shift compensation of preoperative data. Acta Neurochir (Wien)
shifting of white matter tracts during neurosurgical procedures 149:365 378
initial experience. Radiology 234:218 225 61. Roberts DW, Miga MI, Hartov A, Eisner S, Lemery JM,
44. Gasser T, Ganslandt O, Sandalcioglu E, Stolke D, Fahlbusch R, Kennedy FE, Paulsen KD (1999) Intraoperatively updated neuro
Nimsky C (2005) Intraoperative functional MRI: implementation imaging using brain modeling and sparse data. Neurosurgery
and preliminary experience. Neuroimage 26:685 693 45:1199 1207
45. Stadlbauer A, Moser E, Gruber S, Buslei R, Nimsky C, Fahlbusch R, 62. Cao A, Thompson RC, Dumpuri P, Dawant BM, Galloway RL,
Ganslandt O (2004) Improved delineation of brain tumors: an Ding S, Miga MI (2008) Laser range scanning for image guided
automated method for segmentation based on pathologic neurosurgery: investigation of image to physical space registra
changes of 1H MRSI metabolites in gliomas. Neuroimage tions. Med Phys 35:1593 1605
23:454 461 63. Ding S, Miga MI, Thompson RC, Dumpuri P, Cao A, Dawant BM
46. Hastreiter P, Rezk Salama C, Nimsky C, Lürig C, Greiner G, Ertl T (2007) Estimation of intra operative brain shift using a tracked
(2000) Registration techniques for the analysis of the brain shift in laser range scanner. Conf Proc IEEE Eng Med Biol Soc 2007:
neurosurgery. Comput Graph 24:385 389 848 851
Multimodality Imaging Suite: Neo-Futuristic Diagnostic Imaging
Operating Suite Marks a Significant Milestone for Innovation
in Medical Technology
Mitsunori Matsumae, Jun Koizumi, Atsushi Tsugu, Go Inoue, Jun Nishiyama, Michitsura Yoshiyama,
Jiro Tominaga, and Hideki Atsumi
Abstract In February 2006, Tokai University Hospital offi- room. Physically, the MRI, CT, and combined operating
cially opened the imaging operation suite, which is the first theater with X-ray angiography components have been
hybrid neurosurgical procedure suite to combine magnetic installed in three separate bays (Fig. 1), separated by a
resonance imaging, computed tomography and angiography sliding door lined with lead and copper shielding for X-ray
with a neurosurgical operating room. Here, we describe the and MR, respectively When the door is closed, each room
concept of the imaging operation suite and the first 4 years’ may be used independently, enabling unsurpassed flexibility
experience using this suite. and economic efficiency.
For procedures that combine use of MRI and angiogra-
Keywords Imaging operation room Interventional procedure phy, the angiography table can be rotated in the direction of
Intraoperative MRI MRI the MRI room. The connecting bridge is then placed
between the MRI table and the angiography table. The
tabletop can then be moved in both directions.
For procedures that combine CT and angiography, the
Introduction angiography table can be rotated in the direction of the CT
room. Then, the CT table and angiography table are joined
The use of magnetic resonance imaging (MRI) in diagnostic directly, and the tabletop can be moved in both directions.
radiology has made great strides over the last 20 years. In During imaging, the shielding door must be closed to elimi-
many hospitals, MRI is no longer used exclusively for diag- nate imaging artifacts.
nostic imaging. It also plays a crucial role during therapy and For procedures that combine surgery and MRI, the
interventional procedures, and has now been introduced into operating table can be placed in the direction of the MRI
an operating suite. At the Tokai University Hospital, Kana- room, and surgery can be performed normally in that posi-
gawa Japan, seamless integration of MRI and X-ray imaging tion. Intraoperative imaging can be performed at any time,
with a high-end operating table has been accomplished in a provided that the usual precautions are taken to ensure
futuristic and highly efficient approach to image-guided sterility. When a surgeon calls for mixed modality imaging,
surgery [1, 2]. the tabletop carrying the patient can be slid in-line through
the open door between the two rooms of the suite, allowing
reproducible imaging with minimal patient movement and
time delay, while the most appropriate modality can be
Method selected at any particular moment during any procedure
(Figs. 2 and 3).
The center room is equipped with an operating room-
The new facility is equipped with ideally-positioned radio- grade, clean air-conditioning system, operating lights, medi-
logical equipment (computed tomography (CT), MRI, and cal gas outlets, and an in-hospital information system; this
angiography) combined with a fully-functional operating enables various surgical procedures, including craniotomy,
to be undertaken. The adjacent CT and MRI rooms are also
M. Matsumae (*), J. Koizumi, A. Tsugu, G. Inoue, J. Nishiyama, equipped with medical gas outlets, an OR-grade, clean air-
M. Yoshiyama, J. Tominaga, and H. Atsumi conditioning system, and operating lights, etc., for the safe
Department of Neurosurgery, Tokai University School of Medicine,
143 Shimokasuya, Isehara, Kanagawa, 259 1193, Japan transfer of an anesthetized patient to the CT or MRI for
e mail: mike@is.icc.u tokai.ac.jp diagnosis.
216 M. Matsumae et al.
Fig. 1 Schematic diagram of Tokai University’s imaging operating suite. Left: MR scanner, center: angiography equipment combined with
operating room, right: CT scanner. Electric sliding doors enable either combined or single use of the separate areas
Fig. 2 View from the CT

through the angiography with
operating room to the MRI
scanner of the imaging operating
suite at Tokai University Hospital
All ferromagnetic instruments are removed from the sur- Results

gical field, and the wound is covered with a sterile
drape. The patient can, in fact, be moved into and out of One or two cases of neurosurgery and two cases of radio-
the scanner in some 10 20 s, but due account must be taken logical intervention are currently performed per week. In
of the time needed to remove the non-MRI compatible addition, routine diagnosis of a variety of diseases is done
instruments and to adjust the surgical drapes. The results of when this suite is not being used for interventional proce-
the imaging can determine the surgical approach by showing dures. An average of 40 diagnostic CT cases, 16 diagnostic
the presence of any residual tumor that can be safely MRI cases, and one angiography case are performed each
removed, often in areas that cannot be directly observed by day using these facilities, both day and night, and also on
the surgeon. After the intraoperative imaging, the patient is weekends.
returned to the primary surgical position, and resection The indications for using this suite for neurosurgical
continues. procedures include any type of supratentorial or infraten-
Multimodality Imaging Suite 217
Fig. 3 Interventional radiologist

brings his patient into the MR
scanner from the angiography bay
torial glioma, and pituitary adenomas that extend superiorly other indications, such as the evaluation of laser ablation,
and laterally. MRI scans have been taken an average of percutaneous biopsy, abscess drainage, etc., should be in-
1.2 times per procedure, with a maximum of 5 MR imaging cluded for this state-of-art this system in the future.
sessions per surgical procedure. Most MRI scans were per-
formed in the final stage of the surgical procedure to identify
small residual tumor. The choice of MRI scan sequence
depends on the type of tumor. Discussion
For interventional radiology procedures, this suite is used
for the treatment of stroke, intra-abdominal tumors, and any The present imaging operating suite represents the state
vascular occlusive diseases. For example, in patients with of the art in the respective imaging modalities. Together
uterine cervical cancer, chemoinfusion therapy via bilateral they offer exciting opportunities to explore new fields of
internal iliac arteries is performed. To increase drug concen- interactive and interventional imaging. In the field of intra-
tration in the cancer and decrease the infusion of drug into operative MRI, in 1993 Brigham and Women’s Hospital
the extrapelvic organs, balloon-occluded arterial infusion introduced the world’s first intraoperative MRI system, the
therapy is routinely attempted. However, the actual perfu- so-called double doughnut [3]. At Brigham and Women’s
sion depends on each patient’s anatomy, though angiogra- Hospital, they brought MRI into the operating theater. In the
phy does not adequately demonstrate it. The comparison of German city of Erlangen, there is a unique intraoperative
direct perfusion images during arterial infusion of contrast MRI system, the so-called ‘‘twin operating theater’’, in
media with and without balloon-occlusion is helpful to which an independent MRI scanner and surgical unit are
obtain the optimal distribution of anticancer drugs. For housed in two separate rooms [4]. There are several intrao-
patients with portal hypertension or splenomegaly, partial perative MRI systems currently available in the world, in
splenic embolization is often performed. Before partial which the MRI machine is placed in the operating room
splenic embolization, the gastric or pancreatic branches [5, 6], the MRI machine is placed next to the operating
from the splenic artery should be excluded from the emboli- room [2], a small MRI machine is installed under the
zation, and CT during splenic arteriography helps discrimi- operating table [7 9], or a mobile MRI scanner is moved
nate these branches. Because excessive embolization of the into the operating room [10]. With respect to the MRI
splenic artery may induce abscess formation, 60 75% of scanners, there are MRI scanners specially designed for
embolization is thought to be optimal. However, the evalua- intraoperative MR scanning [3, 7 9], and some are routine
tion of necrotic volume is not easy on angiography or even diagnostic MRI scanners modified for use as intraoperative
CT, because congestion of contrast media during emboliza- MRI scanners [2, 5]; the MR scanner field strength started
tion disturbs the images. MRI is useful to demonstrate the low and became relatively high, now reaching 3 T [11].
ischemic foci immediately after the procedure on blood oxygen In the present imaging operating suite, special attention
level-dependent images without any contrast media. Several has been given to economic issues. Our key phrase is
‘‘sharing imaging equipment’’. The center room is the core MRI, MR angiography and venography, and chemical shift
of the suite; it contains angiographic equipment within a imaging. We are planning to introduce thermoablation ther-
fully functional operating room. The tabletop enables both apy for removal of brain tumors under the guidance of MR
imaging and surgical procedures. The most significant fea- temperature mapping. Moreover, we are exploring the pos-
ture of this system is the ease of patient transfer between sibility of applying this imaging operating suite to other
imaging during treatment, e.g., from angiographic examina- fields, such as otorhinology and orthopedics. There are
tion to MRI or CT without changing beds. The patient can be many possibilities for this unique interventional suite in
transferred easily from the operating table to the MRI, and various medical fields.
back to imaging once again. Patient transfer during treat-
ment is performed with the utmost care under the direction Conflict of interest statement We declare that we have no conflict of
interest.
of a specially trained safety supervisor.
Each of the imaging modalities is capable of performing
diagnostic examinations; however, once treatment (for exam-
ple, surgery) commences in the operating room located Reference
in the center of the suite, imaging is easily performed as
required to provide intraoperative information using high- 1. Matsumae M, Fukuyama H, Osada T, Baba T, Mizokami Y,
performance diagnostic equipment. Atsumi H, Ishizaka H, Tsugu A, Tominaga J, Shiramizu H,
Neurosurgical procedures aim to accomplish safe treat- Shimoda M (2008) Fully functional MR compatible flexible
operating table resolves the neurosurgeon’s dilemma over use of
ment with high precision in a short period of time. Here, the intraoperative MRI. Tokai J Exp Clin Med 33:57 60
key that governs the progress of the operation is the intra- 2. Matsumae M, Koizumi J, Fukuyama H, Ishizaka H, Mizokami Y,
operative identification of the extent of preoperative planning Baba T, Atsumi H, Tsugu A, Oda S, Tanaka Y, Osada T, Imai M,
achieved at a given time. We consider that this suite enables Ishigura T, Yamamoto M, Tominaga J, Shimoda M, Imai Y (2007)
World’s first hybrid interventional procedure suite; MRI/X ray/
precise patient treatment within a shorter time period than
Operation suite (MRXO) marks a significant milestone in improve
previously possible because the progress of the operation can ment of neurosurgical diagnosis and treatment. J Neurosurg
now be identified using intraoperative imaging as required. In 107:266 273
practical terms, the system facilitates intraoperative patient 3. Black PM, Moriarty T, Alexander E III, Stieg P, Woodard EJ,
transfer from the treatment room to the MRI examination
room for verification of the patient’s condition and the degree netic resonance imaging and its neurosurgical applications. Neuro
of procedural progress before returning to the treatment room surgery 41:831 845
and return transfer to the examination room for final verifica- 4. Steinmeier R, Fahlbusch R, Ganslandt O, Nimsky C, Buchfelder M,
Kaus M, Heigl T, Lenz G, Kuth R, Huk W (1998) Intraoperative
tion before the operation is completed; surgery proceeds more
magnetic resonance imaging with the magnetom open scanner:
smoothly and safely than previously possible. concepts, neurosurgical indications, and procedures: a preliminary
Operating rooms equipped with a CT scanner, MRI, or report. Neurosurgery 43:739 747
angiographic equipment have been developed previously; 5. Hall WA, Liu H, Martin AJ, Pozza CH, Maxwell RE, Truwit CL
(2000) Safety, efficacy, and functionality of high field strength
however, this newly-built imaging operating suite at Tokai
interventional magnetic resonance imaging for neurosurgery. Neuro
University Hospital is a more evolved, integrated system of surgery 46:632 642
radiological diagnosis/treatment and surgery that represents 6. Nimsky C, Ganslandt O, Von Keller B, Romstöck J, Fahlbusch R
a ‘‘neo-futuristic diagnostic imaging operating suite’’. Tokai (2004) Intraoperative high field strength MR imaging: implemen
tation and experience in 200 patients. Radiology 233:67 78
University Hospital will endeavor to enhance the efficacy of 7. Ahn JY, Jung JY, Kim J, Lee KS, Kim SH (2008) How to over
treatment by actively using this imaging operating suite. come the limitations to determine the resection margin of pituitary
This multimodality imaging intervention suite was deve- tumours with low field intra operative MRI during trans sphenoidal
loped successfully and is another milestone for innovation surgery: usefulness of Gadolinium soaked cotton pledgets. Acta
Neurochir 150:763 771
in medical technology, which reflects the flexible attitude of 8. Hadani M, Spiegelman R, Feldman Z, Berkenstadt H, Ram Z
our hospital with respect to the needs of all kinds of patients. (2001) Novel, compact, intraoperative magnetic resonance imag
This imaging operating suite is not currently used to ing guided system for conventional neurosurgical operating rooms.
perform intra-operative angiography. However, it is likely Neurosurgery 48:799 809
9. Schulder M, Catrambone J, Carmel PW (2005) Intraoperative
that there will be a need to perform intraoperative angio- magnetic resonance imaging at 0.12 T: is it enough? Neurosurg
graphy (e.g., balloon-assisted aneurysm clipping, identifica- Clin N Am 16:143 154
tion of perfusion during removal of an arterio-venous 10. Sutherland GR, Kaibara T, Louw D, Hoult DI, Tomanek B,
malformation nidus) using this suite in the near future. The Saunders J (1999) (1999) A mobile high field magnetic resonance
system for neurosurgery. J Neurosurg 91:804 813
intraoperative functional techniques that have so far been 11. Pamir MN, Peker S, Ozek MM, Dinçer A (2006) Intraoperative
used for surgical decision-making with this imaging MR imaging: preliminary results with 3 Tesla MR system. Acta
operating suite have been MR spectroscopy, functional Neurochir Suppl 98:97 100
Improving Patient Safety in the Intra-operative MRI Suite Using
an On-Duty Safety Nurse, Safety Manual and Checklist
Mitsunori Matsumae, Yasuhiro Nakajima, Eiji Morikawa, Jun Nishiyama, Hideki Atsumi, Jiro Tominaga,
Atsushi Tsugu, and Isao Kenmochi
Abstract This paper describes the use of an on-duty safety procedure, and the surgical crew is not as accustomed to this
nurse, a surgical safety manual and a checklist as an essential kind of special procedure. There are many steps required to
precursor to evaluating how these approaches affect surgical take an MR image during surgery. Additionally, teamwork
quality, communication in surgery crews and contribute to in the intra-operative MRI suite is an important component
the safety of surgical care in the intra-operative magnetic of operation room efficiency, quality of surgical care, and
resonance imaging (MRI) suite. patient safety. For this reason, the surgical safety manual and
checklist and the on-duty safety nurse are needed to manage
Keywords Checklist Intra-operative magnetic resonance concerns or ambiguities related to procedures, promote team
imaging Manual On-duty safety nurse Surgical safety building, and keep the surgical workflow smooth.
Introduction Materials and Methods
The use of intra-operative imaging is gradually increasing in

On-Duty Safety Nurse
the field of neurosurgery. In February 2006, our hospital
officially opened the Magnetic Resonance/X-ray/Operation The aim of the on-duty safety nurse is to focus only on safety
Suite (MRXO) which is a novel surgical suite that includes issues. The on-duty safety nurse functions independently
an interventional radiology system for intra-operative MRI. from nurses involved in the surgical procedure, and behaves
This imaging operation suite represents a major advance in as a third person in the imaging operation room. During
the field of neurosurgery, and will facilitate the development preparation for the intra-operative MRI, the on-duty safety
of many new clinical techniques. The details of this imaging nurse reads aloud a surgical safety manual and checklist
operation suite were reported previously [1, 2]. (Fig. 1). The on-duty safety nurse is requested to speak up
We introduced an on-duty safety nurse, a surgical safety and provide clear instructions during the surgery. The surgi-
manual and a checklist in this imaging operation suite. There cal safety checklist is read aloud by the on-duty safety nurse
are several reasons for introducing this safety system into and requires oral confirmation by all other crew in the intra-
this imaging operation suite. During interventional proce- operative suite. All crew in the operation room, including the
dure specially designed for the intra-operative MRI, the surgeon, anesthesiologist, scrub and circulation nurses, and
surgical crew must pay special attention to both surgical radiology technician, are required to follow the on-duty
and MRI safety. Intra-operative MRI is not a routine surgical safety nurse’s instructions.
M. Matsumae (*), J. Nishiyama, H. Atsumi, J. Tominaga, and

A. Tsugu Surgical Safety Manual
Department of Neurosurgery, Tokai University School of
Medicine, 143 Shimokasuya, Isehara, Kanagawa 259 1193, Japan
e mail: mike@is.icc.u tokai.ac.jp
The surgical safety manual defines six phases of the surgery:
preoperative inspection (42 items), before skin incision
Y. Nakajima, E. Morikawa, and I. Kenmochi
Division of Emergency and Critical Care Unit, Tokai University (16 items), preparation for the intra-operative MRI (75
Hospital, 143 Shimokasuya, Isehara, Kanagawa 259 1193, Japan items), after taking the intra-operative MRI (53 items),
The phase of preoperative inspection includes: confirma-

tion of the conditions of all surgical and imaging facilities,
identification of the surgical procedure, verification of the
patient’s name, and introduction of the crew to each other.
The phase of before skin incision includes: a clearance
check for the MRI bore (Fig. 2), and verification of whether
the right procedure is being performed on the right patient at
the right head site.
The preparation for the intra-operative MRI phase has
many points. For intra-operative MRI, surgery can be per-
formed in the usual fashion with usual surgical instruments
(not MRI compatible). Intra-operative imaging can be per-
formed at any time. Once the surgeon wants to take the intra-
operative MRI, the on-duty safety nurse takes control of the
procedure. The on-duty safety nurse instructs the surgeons and
scrub nurse to remove all ferromagnetic instruments from the
surgical field, and the surgical wound is covered with a sterile
drape while the tabletop is placed in a flat position. The
circulation nurse places the patient transfer system between
the operating table and the MRI table on the order of the
on-duty safety nurse. Once the on-duty safety nurse checks
Fig. 1 The safety checklist is orally implemented by the on duty safety
nurse with oral confirmation by other crew members. The on duty the patient’s transfer system, the locking system that unlocks
safety nurse improves team communication and enables smooth work the tabletop from the operation table base can be released.
flow during surgery After the safety checklist is completed, the on-duty safety
nurse allows the patient to be moved into the MRI scanner.
after the operation (20 items), and non-normal (emergency) The phase of after taking intra-operative MRI includes:
condition (32 items). The surgical safety manual and check- replacing the patient transfer system between the MRI table
list identifies each phase corresponding to a specific period and operation table; the on-duty safety nurse repeats the
in the workflow of the intra-operative MRI. In each phase, patient moving step in reverse order of the phase of prepa-
the surgical safety manual and checklist help surgical crews ration for the intra-operative MRI process.
confirm that the critical safety steps are completed before The phase of after surgery includes: confirming the
proceeding with the operation and other activities. patient’s vital signs and verifying the type of procedure,
Fig. 2 After the patient position

is set, and the head is secured and
fixed, both the neurosurgeon and
on duty safety nurse check the
clearance between the patient’s
body and the MRI bore to avoid a
collision. The on duty safety
nurse is applying the dummy MRI
bore above the patient. The
surgeon makes sure the patient
position allows an approach into
the MRI bore before starting the
operation
Improving Patient Safety in the Intra-operative MRI Suite Using an On-Duty Safety Nurse, Safety Manual and Checklist 221
specimen labels, instrument counts, and liaison points from surgical safety manual in the last 3 years. It was noted that
operation room to intensive care unit. the learning curve over the past 3 years showed that the time
Non-normal events may include: emergency stop of the required for the safety protocol has gradually shortened by
MRI, emergency opening of the shielded door, alerts from the about 8 min. There were no significant differences in time
oxygen monitor and vital sign checks, handling problems with required for the safety protocol and in the workflow pattern
the patient transfer system, shielded door, or the operation for each on-duty safety nurse.
table, and handling the emergency communication system. The on-duty safety nurse behaved as a sophisticated crew
leader, and maintained workflow during the surgery. The
surgical safety manual eradicated ambiguous working pro-
Surgical Safety Check List cesses for intra-operative MRI. We successfully developed
good communication and a well organized intra-operative
MRI crew using this safety tool.
At each workflow key point, the surgical safety checklist
provides specific checkpoints to be verified by the surgical
on-duty safety nurse. The specific checkpoints in the work-
flow are: phase of preoperative inspection (4 points), before Discussion
skin incision (4 points), preparation for the intra-operative
MRI (10 points) (Table 1), after taking intra-operative MRI Neurosurgical care has seen rapidly increasing technology
(6 points), after surgery concluded (4 points), and non- advances worldwide for more than a century, including the
normal (4 points). introduction of a navigation system and an imaging operation
system. Today, we can easily introduce new technology into
our operation room. However, introducing a new technology
Results may drive a patient into an unsteady condition. To prevent
this situation, we needed a new tool which works well in the
In the present series of patients, there was no incident or field of patient safety, especially surgical safety.
accident during patient transport or intra-operative imaging Not only surgeons, but all humans are fallible; therefore,
in the 3 years since the intra-operative imaging operation relying solely on memory to carry out all tasks may cause
suite opened. The time required for the phase of preparation people to commit errors that could result in adverse events
for the intra-operative MRI process was an average of [3]. There are many causes for human errors, including
22 min, and the phase of after taking intra-operative image several psychological factors such as anxiety, ambiguous-
was 12 min on average. The intra-operative MRI required an ness, impatience, fear, interruptions, and time stress [4, 5].
average of 13 min. The total extra time required for com- These psychological factors should be eliminated in all crew
pleting the safety manual and intra-operative MRI averaged members in the operation room. Teamwork in the operation
47 min. There has been some minor modification of the room is an important component of operation room effi-
ciency, and the quality of surgical care and patient safety is
Table 1 Elements of the surgical safety checklist generally emphasized [6, 7]. Careless and inadvertent mis-
Checklist for preparation to the intraoperative MRI takes usually occur in routine workflow. It is important to
Check points Response Responder verbalize instructions using simple and standardized words,
All ferromagnetic materials Removed CN especially when the speaker and listener are physically
removed from the operation separated [8, 9]. Surgical crews require explicit communica-
table tion with each other, and should understand the role of
Surgical instruments count Completed SN standardization in effective team functioning as it relates to
Bed position Flat CN
policies and procedures [9]. The scientific literature indi-
MRI bore clearance Clear NS
Surgical field covered by Covered NS cates that adverse events can be avoided if proper tools are
sterilized cap available in working environments [9]. Process manuals and
Shielded door open Opened RT checklists are some of the tools that have been successful in
Patient transfer system Clear RT other high-risk industries, such as airline cockpits and nuclear
Anyone dose not have any Clear All crew power plants, as well as in the medical field [10, 11].
ferromagnetic materials
In 2008, the World Health Organization launched a new
Vital signs check Normal AN
Safety lock Unlocked CN safety checklist for surgical teams to use in operating theatres
Check list completed Cleared for transfer SON as part of a major drive to make surgery safer around the world
CN circulaqtion nurse, SN scrub nurse, NS neurosurgeon, RT radiation [12]. This safety checklist worked 19 points into a simple one-
technician, AN anesthesiologist, SON safety on duty nurse page checklist developed in consultation with international
experts in surgery, anesthesiology, nursing and patient safety, World’s first magnetic resonance imaging/x ray/operating room
and divides surgical procedures into three phases. This surgi- suite: a significant milestone in the improvement of neurosurgical
diagnosis and treatment. J Neurosurg 107:266 273
cal safety checklist was used in eight hospitals in eight cities 2. Matsumae M, Fukuyama H, Osada T, Baba T, Mizokami Y,
around the world for 1 year. The results of this study showed Atsumi H, Ishizaka H, Tsugu A, Tominaga J, Shiramizu H,
that a checklist-based program was associated with a signifi- Shimoda M (2008) Fully functional MR compatible flexible
cant decline in the rate of complications and death from operating table resolves the neurosurgeon’s dilemma over use of
intraoperative MRI. Tokai J Exp Clin Med 33:57 60
surgery in a diverse group of institutions around the world 3. Leape LL (1994) Error in medicine. JAMA 272:1851 1857
[13]. In each region and country, institutions can modify this 4. Gibbs VC (2005) Patient safety practices in the operating room:
list to fit each type of surgical procedure. Our intra-operative correct site surgery and nothing left behind. Surg Clin North Am
MRI team introduced a surgical safety checklist in our imag- 85:1307 1319
5. Perazzelli M (2007) Operating room working procedures: a good
ing operation suite. Our safety surgical checklist has many tool for patient safety? Perioper Nurs Clin 2:255 257
checkpoints when compared with that of the World Health 6. Makary MA, Sexton JB, Freischlag JA, Holzmueller CG, Millman EA,
Organization, but the concept and aim are similar. In the 3 Rowen J, Operating Room Teamwork among Physicians and
years since the opening of our imaging operation suite, we Nurses (2006) Teamwork in the eye of the beholder. J Am Coll
Surg 202:746 752
have not had an incident or accident related to items in the 7. Sexton JB, Makary MA, Tersigni AR, Pryor D, Hendrich A,
surgical safety manual and checklist. Thomas EJ (2006) Teamwork in the operating room: frontline
While intra-operative MRI is a relatively new technique, perspectives among hospitals and operating room personnel.
its use has gradually increased in the neurosurgical field over Anesthesiology 105:877 884
8. Pronovost P, Berenholtz S, Dorman T, Lipsett PA, Simmonds T
the past 15 years [14], even though it requires extra opera- (2003) Improving communication in the ICU using daily goals.
tion time to perform MRI. When introducing this kind of J Crit Care 18:71 75
new imaging technique into surgery, there are many process- 9. Sexton BJ, Grote G, Naef W, Straeter O, Helmreich RL (eds)
es needed to keep the workflow smooth. This is the reason (2004) The better the team the safer the world: golden rules of
group interaction in high risk environments: evidence based sug
we introduced the on-duty safety nurse and the surgical gestions for improving performance. Gottlieb Daimler and Karl
safety manual into our imaging operation room. Our on- Benz Foundation, Ladenburg, pp 1 58
duty safety nurse plays a very powerful role as a third 10. Lingard L, Espin S, Rubin B, Whyte S, Colmenares M (2005)
person. The surgical safety manual is implemented by the Getting teams to talk: development and pilot implementation of a
checklist to promote interprofessional communication in the OR.
on-duty safety nurse step by step, and the surgical safety Qual Saf Health Care 14:340 346
checklist is called by the on-duty safety nurse at key points 11. Lingard L, Regehr G, Orser B, Reznick R, Baker GR, Doran D,
in the process. The surgical safety manual provides a defined Espin S, Bohnen J, Whyte S (2008) Evaluation of a preoperative
set of steps to remove ambiguities in work processes. These checklist and team briefing among surgeons, nurses, and anesthe
siologists to reduce failures in communication. Arch Surg 143:
new approaches lead to sustainable team building, provide 12 17
education for new surgical crew staff, and keep the imaging 12. World Health Organization World Alliance for Patient Safety
surgical procedure workflow smooth. (2009) Implementation manual surgical safety checklist, 1st edn.
http://www.who.int/patientsafety/safesurgery/tools resources/
Conflict of interest statement We declare that we have no conflict SSSL Manual finalJun08.pdf. Accessed 21 Sept 2009
of interest. 13. Haynes AB, Weiser TG, Berry WR, Breizat LSR, AS DEP,
Herbosa T, Joseph S, Kibatala PL, Lapitan MCM, Merry AF,
Moorthy K, Reznick RK, Taylor B, Gawande AA (2009) Surgical
safety checklist to reduce morbidity and mortality in a global
population. N Engl J Med 360:491 499
References 14. Black PM, Moriarty T, Alexander E III, Stieg P, Woodard EJ,
1. Matsumae M, Koizumi J, Fukuyama H, Ishizaka H, Mizokami Y,
netic resonance imaging and its neurosurgical applications. Neuro
Baba T, Atsumi H, Tsugu A, Oda S, Tanaka Y, Osada T, Imai M,
surgery 41:831 845
Ishiguro T, Yamamoto M, Tominaga J, Shimoda M, Imai Y (2007)
Operating Room Integration and Telehealth
Richard D. Bucholz, Keith A. Laycock, and Leslie McDurmont
Abstract The increasing use of advanced automated and to help control a movement disorder such as Parkinson’s
computer-controlled systems and devices in surgical proce- disease. In such cases, the location of the electrode is dictated
dures has resulted in problems arising from the crowding by the patient anatomy and data from intraoperative recording
of the operating room with equipment and the incompatible of brain activity using microelectrodes. Once the device
control and communication standards associated with each has been placed, its position must then be confirmed by
system. This lack of compatibility between systems and recordings from the electrode and intraoperative imaging.
centralized control means that the surgeon is frequently The first step in planning such a procedure is to obtain
required to interact with multiple computer interfaces in image data of the target in order to determine its exact
order to obtain updates and exert control over the various location. Using scan data acquired in multiple planes, a
devices at his disposal. To reduce this complexity and pro- virtual 3D model of the patient’s neuroanatomy can be
vide the surgeon with more complete and precise control of constructed using a computer navigation system such as
the operating room systems, a unified interface and commu- the Medtronic StealthStation, allowing the optimum path to
nication network has been developed. In addition to improv- the target to be determined.
ing efficiency, this network also allows the surgeon to grant Before surgery begins, the 3D space of the navigation
remote access to consultants and observers at other institu- system is registered with the patient anatomy to ensure
tions, enabling experts to participate in the procedure with- precise localization of the target and surgical instruments.
out having to travel to the site. The surgical procedure then commences with the insertion
of the electrode into the skull, followed by its advancement
Keywords Integration Interface Remote access Surgical to the target under guidance from the navigation system.
communication network SurgON Though the preoperative planning ensures accurate position-
ing of the electrode with respect to the target, it is necessary
to fine-tune the position of the electrode by reference to
electrophysiological recordings acquired during the proce-
Introduction dure itself. Once the optimum position is determined, the
recording microelectrode is withdrawn and replaced with the
stimulation electrode, whose position is then verified with
In recent decades, the introduction of computer-based imag-
further imaging.
ing and navigation systems has revolutionized many aspects
A neurosurgical procedure of this type thus involves
of neurosurgery, resulting in improved precision and control
(at least) six distinct steps:
and enhanced ability to make adjustments to the surgical
plan during the course of a procedure. • Imaging
A representative neurosurgical procedure that benefits • Planning
from these advances is the placement of a deep brain electrode • Navigation
• Recording
• Therapy (the stimulation electrode)
R.D. Bucholz (*), K.A. Laycock, and L. McDurmont • Confirmation by imaging
Division of Neurosurgery, Department of Surgery, Saint Louis
Each of these steps involved a computer, networking and
University Health Sciences Center, 3635 Vista Avenue, Saint Louis,
MO 63110, USA clinician interaction. However, current operating rooms are
e mail: Richard@bucholz.org poorly designed to support such complex procedures, while
224 R.D. Bucholz et al.
personnel and economic constraints further interfere with dissociated for use elsewhere. Integration of a device within
optimal performance. a given procedure requires interactions between devices in
An obvious problem is that the various systems and terms of control, data (video, audio) and automation (with
devices in the operating room (OR) are developed and man- respect to positional information). Furthermore, the afore-
ufactured by unrelated companies and are not designed to mentioned turnover in OR personnel dictates that control
interact efficiently. On the personnel front, the perennial interfaces must be relatively simple. To facilitate this, a
shortage of trained staff has been exacerbated in recent standard for communication is clearly required [2].
years by economic considerations. OR staff must be able to The rise of a similar problem in the field of radiology
support a variety of cases staff specialization in one partic- resulted in the development of the DICOM standard. This
ular field of surgery is not practical at many hospitals and allowed hospitals to connect multiple different diagnostic
this requires appropriate levels of training and experience. and treatment devices and have them interact and share
Unfortunately, having acquired the requisite experience, information. Unfortunately, the standard is constantly evol-
they tend to seek more attractive appointments, resulting in ving, with various add-on specialties (such as dosimetry and
constant turnover. Likewise, experienced neurophysiologists image guidance) having being introduced after the initial
are also hard to find and may have to be brought in from standards had been established. Furthermore, it has been
other institutions to assist with certain cases, often at signifi- discovered that many devices theoretically capable of
cant cost. being integrated through DICOM actually communicate
Just as the OR staff must be able to support various types in different ‘‘dialects’’, thus impairing the reliability and
of procedures, so the ORs themselves are expected to be practicality of communication.
able to accommodate any kind of case. Few hospitals can Nevertheless, despite these limitations, it is time to intro-
dedicate ORs to a single purpose; an open-heart procedure duce a similar standard for communication between and
may follow a craniotomy case in the very same OR. Need- control of surgical devices. Such a system would greatly
less to say, with such a bewildering variety of procedures relieve the technological chaos that currently prevails in
being performed in rapid succession in a given location, the OR, and if properly designed would be able to satisfy
often involving many of the same support staff, a significant the needs of all users, regardless of their particular surgical
proportion of medical errors occur in the OR. specialty. The result would be less expensive, more efficient,
Despite the variety of procedures which may be sched- and higher quality medical care.
uled for a given OR, there are certain requirements common
to all surgical interventions:
• Shielding (against microbial contamination and to protect
information) An Integrated Surgical Communication
• Documentation (in the form of audio, video and still Network: SurgON
image records)
• Life support
Our group has developed a prototype of an integrated surgi-
• Visualization
cal communication network as outlined above in the form of
• Therapeutics
SurgON (SURGical Operative Network) [3]. In this system,
• Plasticity
each device acts as a web server, storing the web pages that
Devices and systems that are required for most surgical serve as control panels for that particular device. These web
procedures can be left in place in the OR, but when addi- pages are used locally to control the devices, each of which
tional equipment is introduced for particular procedures, this has an Ethernet port. An IP address is assigned upon attach-
can result in crowding and obstruction of the OR staff. ing a device to the SurgON local area network (LAN), and
Furthermore, the various incompatible systems require the the appropriate web page is transmitted as requested. The
surgeon and his team to monitor and interact with multiple surgeon’s controller allows secure routing of data and con-
control interfaces and displays, impairing their efficiency trol streams via a firewall to the hospital and beyond.
and increasing the risk of unforeseen complications. Conse- For such an approach to be effective, several design
quently, there is intense pressure to integrate devices, both to objectives must be met. First, each device must be assured
reduce the number of visualization and control displays and of a specific bandwidth, and every type of surgical device
to permit more effective control [1]. must be controllable locally, even in the event of complete
As it is, the plasticity required of a modern OR places network failure. It is also desirable that devices can be con-
tremendous stress on the system. Complex devices must be nected and removed with instant recognition, while data
easy to move, capable of instant integration with the systems streams can be shared over the Internet using a graphic
already in use in the OR, robust and fault-tolerant, and easily user interface.
Operating Room Integration and Telehealth 225
With respect to the control of devices, three classes of

control have been identified:
• Simple control (request-based control and telemetry)
• Compound control (synchronized control and telemetry;
shareable bidirectional datastreams)
• Complex control (synchronized control and feedback to
multiple users; automated interaction with other devices)
When a device to be controlled is plugged in, it becomes
available on the OR network through its web page. Alterna-
tively, radio frequency shielding in the OR will allow a
wireless device to be recognized as it is brought into the
shielded room. The web control page is transmitted to con-
trol devices when selected, and the device becomes control-
lable locally or remotely (as permitted). Fig. 1 Schematic layout for a network controlling multiple devices
from the surgeon’s console
This approach offers several benefits, the most obvious of
which is the ability to use one cable connection for multiple
data streams. The surgeon can monitor any data source remote consultants could also be represented as icons, with
during a procedure, and the video stream is also viewable their observation or participation being enabled by a drag-
by nursing and anesthesia staff. Control of complex diagnos- and-drop interface. Figure 2 shows a screenshot of the sur-
tic equipment would be simplified, and the surgeon’s own geon’s console as configured to control a Möller VM900
control preferences could be distributed in an automated microscope, a bipolar coagulator and a Tetrad ultrasound
fashion over many devices. For optimum efficiency, each imaging system. A surgeon’s particular preferences can be
device could be automatically aligned to a chosen surgical stored on appropriate media and uploaded prior to surgery.
path, and a high dedicated bandwidth would ensure precise The single unified multi-source display system would
control. If desired, one or more other control systems could avoid the need for multiple computer monitors in the OR.
be implemented as a separate surgical controller (voice, The display device could take one of several forms as long as
gesture, etc.), and a remote assistant or expert can be given it is compatible with a sterile cover for use in the OR. Our
access to the systems as needed [4]. current test model uses an LCD panel on wheels, but wall-
Figure 1 shows a schematic layout for a network mounted plasma screens or head-mounted displays would be
controlling multiple devices from the surgeon’s console. equally practical, and even a personal data assistant (PDA)
This console would display a dynamic list of available could be used.
devices, each represented by an icon. Clicking on the icon Devices producing information would have a streaming
for a particular device would bring up its controls. Similarly, web server; pertinent information would be digitalized and
Fig. 2 Screenshot of the

surgeon’s console as configured
to control a Möller VM900
microscope, a bipolar coagulator
and a Tetrad ultrasound imaging
system
226 R.D. Bucholz et al.
Fig. 3 Representative screen shot of a remote station displaying control interfaces for a Möller VM900 microscope and Tetrad ultrasound system
sent out as a data stream. Thus, a microscope would produce standard Internet software (Internet Explorer, Java plug-in,
a video stream, the life-support equipment would generate NetMeeting, etc.) should be adequate. Thus, almost any
EKG readouts, and the neurophysiological monitor would current computer is capable of being used as a remote
produce an audio stream. controller.
Every control panel and information stream can be shared
outside the operating room as required. To facilitate such
remote access, the dynamic firewall software provides a Conclusion
graphic interface that allows sharing of information with a
remote consultant equipped with nothing but a PC. It isolates
It is our belief that SurgON addresses the needs of surgery of
local network traffic from the wide area network (WAN) and
the future. It provides the requisite shielded space, empow-
guarantees bandwidth to the surgeon. The local surgeon
ering the surgeon to control all informational streams. It
retains complete control over external connections into the
fulfills life support requirements through its ability to be
OR: upon logging in, a remote consultant only sees the
networked and displayed on request to all clinicians
locations of surgeons currently using the system. Controls
involved with a case, while the associated display devices
could also be shared to maximize device optimization
allow seamless communication between the clinicians.
(image quality, etc.).
There is a protocol for instant recognition of all devices
The local surgeon can grant outside viewing or control as
entering the protected space, and the system provides the
needed. By dragging the selected device icon over the
option to share control of intraoperative devices with remote
remote consultant’s icon, a logical connection is established.
consultants and assistants.
The local surgeon can also stipulate the type of access to
each device, e.g., data sharing, joint control or autonomous Conflict of interest statement We declare that we have no conflict
control by the remote surgeon. The control and/or connec- of interest.
tion can also be revoked as needed. Figure 3 shows a repre-
sentative screen shot of a remote station, displaying the
control interfaces for the microscope and the Tetrad ultra- References
sound system.
For the SurgON system to work with a remote work-
1. Healthcare Financial Management Association (2003) Achieving
station, a suitable network connection is required. For operating room efficiency through process integration. Healthc
control signals and audio, a modem is fast enough, and Financ Manage 57(3):Suppl 1 7, following 112
Operating Room Integration and Telehealth 227
2. Gallagher AG, Smith CD (2003) From the operating room of the international conference on medical image computing and comput
present to the operating room of the future. Human factors lessons er assisted intervention (MICCAI 2000), Pittsburgh, Pennsylvania,
learned from the minimally invasive surgery revolution. Semin October 2000
Laparosc Surg 10(3):127 139 4. Feussner H (2003) The operating room of the future: a view from
3. Bucholz RD (2000) SurgON: a standard for networked commu Europe. Semin Laparosc Surg 10(3):149 156
nication between surgical devices (keynote presentation). Third
Other Intraoperative Imaging Technologies
and Operative Robotics
Intra-operative Robotics: NeuroArm
Michael J. Lang, Alexander D. Greer, and Garnette R. Sutherland
Abstract This manuscript describes the development and Meanwhile, human capacity for non-linear processing incor-
ongoing integration of neuroArm, an image-guided MR- porates incomplete data sets and prior experience, and is
compatible robot. necessary for anticipatory action [5]. At the human-machine
Methods: A neurosurgical robotics platform was devel- interface (HMI), the unique attributes of robots and humans
oped, including MR-compatible manipulators, or arms, with may well result in improved surgical intervention and out-
seven degrees of freedom, a main system controller, and come. These advantages are exploited in the master-slave
a human-machine interface. This system was evaluated configuration of most medical robotic devices [6], and
during pre-clinical trials and subsequent clinical application, are why automation, in order to be safe, would require a
combined with intra-operative MRI, at both 1.5 and 3.0 T. fundamental change to processing architecture and software
Results: An MR-compatible surgical robot was success- design.
fully developed and merged with ioMRI at both 1.5 or 3.0 T. Following the modification of an industrial robot
Image-guidance accuracy and microsurgical capability were (Programmable Universal Machine for Assembly (PUMA),
established in pre-clinical trials. Early clinical experience Advance Research and Robotics, Oxford, Connecticut)
demonstrated feasibility and showed the importance of a for neurosurgical use, various robotic systems have been
master-slave configuration. Surgeon-directed manipulator adapted to or created for neurosurgery [2]. These systems
control improved performance and safety. have seen only isolated acceptance, largely due to a combi-
Conclusion: NeuroArm successfully united the precision nation of safety concerns, complexity of use, intra-operative
and accuracy of robotics with the executive decision-making imaging incompatibility, cost, and restricted range of appli-
capability of the surgeon. cation. In 2002, development of an MR-compatible robot
was initiated at the University of Calgary in an effort to
Keywords Human-machine interface Image-guided surgery integrate the precision and accuracy of robotics with the
Microsurgery Robotic surgery Stereotaxy imaging capabilities of an innovative intra-operative MRI
(ioMRI) system based on a moveable magnet [7, 8]. The
master-slave dynamic governing neuroArm is discussed,
together with the neurobiological mechanisms which sepa-
rate human and computer decision-making capability.
Introduction
Robotic technology has been applied to surgery for

many years [1 3]. As with industrial manufacturing applica- Materials and Methods
tions, robotics has demonstrated precision and accuracy that
exceeds human ability, particularly for repetitive tasks [4]. The design and manufacture of neuroArm has been previ-
ously reported [3, 9]. In summary, the system consists of two
MR-compatible manipulators mounted on a mobile base, a
main system controller, and a human-machine interface
(HMI), or workstation. The manipulators, or arms, have
M.J. Lang, A.D. Greer, and G.R. Sutherland (*)
seven degrees of freedom (DOF), and are able to grasp and
Department of Clinical Neurosciences, University of Calgary, Calgary,
AB, Canada manipulate a variety of neurosurgical instruments. The main
e mail: garnette@ucalgary.ca system controller processes the computational needs of
robotic manipulation, and mediates the reciprocal exchange workstation monitors. In this configuration, the manipulator
of information between surgeon and machine. Finally, the is registered to the magnet isocenter.
HMI provides the sensory milieu in which the surgeon is The main system controller governs the transmission of
able to perform surgery within an image-guided environment. input signals to the robot, as well as position and force data
To develop an image-guided robotic system, manipulator returning to the HMI. The software currently being utilized
arms were constructed from MR-compatible materials, such in the main system controller took over four years to deve-
as titanium and polyetheretherketone. Piezoelectric motors lop, given the requirements of image guidance, quantified
were chosen for their MR-compatibility, 20,000-h lifetime, force feedback, and safety. This software not only acts as a
and intrinsic braking characteristics in the event of sudden throughput for data, but actively monitors all aspects of the
power loss. The system includes dual-redundant rotary elec- system. In the event of divergence from intended manipula-
tric encoders, which record joint position to an accuracy of tor motion, the embedded safety mechanisms halt further
0.01 degrees. The end-effector incorporates a unique mech- movement. However, while system software can recognize
anism for tool grasping, roll, and actuation, such as opening positional error, the surgeon may be able to recognize the
or closing bipolar forceps or scissors. In addition, the con- context of unintended movement more rapidly, and thus
figuration maintains sterility via end-effector attachments avoid injury to adjacent neural tissue. This is particularly
which pierce the manipulator drapes while capping the relevant given the speed of surgery, in which small varia-
break in the sterile field (Fig. 1). Two titanium six-axis tions in reaction time can result in substantially different
force/torque sensors in contact with the tool generate force outcomes. To avoid such a critical event, a foot-pedal, used
data in three translational DOF, which is fed back to the as a dead-man switch and wired directly into the process
surgeon via two handcontrollers. For microsurgery, both logic controller, was added to allow the surgeon to rapidly
manipulators are transferred to a mobile base, along with a halt the manipulators in the event of unintended movement.
6-DOF digitizing arm. The digitizing arm is used to register Union of the precision and accuracy of machine technol-
the manipulators to fiducials located within the radiofre- ogy and human executive capacity is made possible by the
quency (RF) coil, and hence to pre- or intra-operatively HMI apparatus [10]. The surgeon is immersed in sensory
acquired MR images. For stereotaxy, one manipulator is data, including 3D intra-operative MR images with tool
transferred to a polyetheretherketone platform attached to overlay, virtual position of the manipulators relative to the
the gradient insert within the bore of the magnet. The plat- RF coil, stereoscopic display of the surgical field from the
form includes two MR-compatible cameras, which allow operating microscope, an image of the operative area, and
visualization of the operative site and manipulator from the recreated sense of touch, which is displayed in Newtons and
Fig. 1 Integration of human and

machine: the figure shows
neuroArm positioned for removal
of an olfactory groove
meningioma. The right arm holds
a bipolar forcep and the left a
bayonet shaped sucker.
Stereoscopic images from two
high definition cameras attached
to the operating microscope are
projected to the surgeon at the
workstation (insert). Also shown
in the insert, are the 3D MRI
display (left) and virtual
manipulators registered to the
RF coil (right)
Intra-operative Robotics: NeuroArm 233
Fig. 2 Three dimensional MRI

display: the image demonstrates a
selected surgical corridor with
tool overlay. Once delineated, the
tools may only pass within the
limits of the pre defined corridor
also pictorially represented. Force feedback, essential for implantation into predetermined targets using a cadaveric
microsurgery, permits quantification of the forces of dissec- model. Accuracy was found to be equal or better to existing
tion and the potential to set explicit limits to applied force. In surgical navigation technology. Unlike traditional naviga-
addition, the surgeon can construct virtual surgical corridors tion technology, location of the implanting device was con-
on the MRI display (Fig. 2). Once established, tool manipu- firmed with real-time imaging prior to nanoparticle release.
lation may occur only within the limits of these Clinical application was initially demonstrated in five
pre-determined pathways. Within this sensory immersive patients with various intracranial neoplasms [9]. NeuroArm
environment, the surgeon is able to safely control the mani- was progressively introduced in a staged manner, so as to
pulators. isolate specific issues relating to its use in the operating
room. Early cases focused on the positioning and draping
of the manipulators for use within the sterile field. Assess-
ment of image-guidance accuracy was then performed, fol-
Results lowed by initial use during neoplasm resection. During the
fifth case, an encoder failure occurred, resulting in uncon-
As previously published, the surgical robotic system was trolled motion causing the suction tool to come into contact
designed, manufactured, and integrated into a neurosurgical with a retractor. This event triggered a safety review, out of
operating room over a 6-year interval [3]. During the which the aforementioned foot-pedal was developed and
subsequent 18 months, deficiencies in both hardware and incorporated. Despite the occurrence of minor technical
software were identified and resolved. The foot-pedal, complications, patient safety remained uncompromised.
added during this interval, was found to increase manipula- Sample size during initial clinical use was limited by the
tor control, further enhancing safety. With usage, errors decision to upgrade the 1.5 T ioMRI system to 3.0 T. This
related to positioning data and image registration were was complicated by the decision to shift from local to whole-
likewise identified and resolved. room RF-shielding. NeuroArm was integrated into this new
Pre-clinical trials were performed using rats and cadavers ioMRI environment. To avoid RF-interference from elec-
[9]. To establish microsurgical application, two qualified tronic cables controlling neuroArm, a penetration panel was
microsurgeons achieved comparable results using either constructed. At the site of shield penetration, RF-noise is
established microsurgical techniques or neuroArm. The test- eliminated with line filtration and copper to fiber-optic con-
ed procedures consisted of splenectomy, nephrectomy, and verters. Conversion to whole-room shielding simplified the
submandibular gland resection in a rodent model. Naviga- method for registration during stereotaxy. Manipulator
tional accuracy was established by means of nanoparticle placement on a polyetheretherketone platform, attached to
Fig. 3 Image guided robotic

assisted stereotaxy at 3.0 T:
this photograph shows the right
manipulator, located on a
platform attached to
the gradient insert within the bore
of the moveable 3.0 T magnet.
Cables are connected to two MR
compatible cameras, allowing
visualization of the manipulator
and surgical field
the gradient insert within the magnet bore, allows image evaluating multiple choice alternatives simultaneously, but
registration to be based on the magnet isocenter (Fig. 3). also continued processing and modification of such choices
after the action has been initiated. Consequences of a deci-
sion are then incorporated as experience into future heuris-
tics. Presently, computers, which do not rely on the
Discussion relatively slow synaptic connections of neuronal transmis-
sion, utilize computational circuitry and software designed
Pre-clinical trials and initial clinical experience with neu- with inherent linearity, although this may change [16]. The
roArm have demonstrated that the system safely unites surgeon controlling neuroArm can evaluate the context of
robotic precision and accuracy with human decision- unintended motion, resulting in a much faster reaction than
making. This connection is based on a master-slave configu- the main system controller.
ration, similar to most surgical robotic systems. Such a In addition to differences at the level of basic compu-
design was chosen largely due to limitations in contempo- tation, higher-level data processing is critical to human
rary computer technology, but was fortuitous nonetheless. executive functioning. Step-by-step deduction in computer
Indeed, the master-slave dynamic is indispensable to the processing requires the evaluation of all possible choices to
practice of robotic surgery, owing to the nature of human conclusion or a pre-determined fail condition, prior to select-
decision-making. While the superior precision and accuracy ing a single option with the highest likelihood of success.
of robotic movement is accepted [2], the mechanisms separ- When applied to surgery, these distinctions have substantial
ating human and computer decision-making capability in the impact as decision complexity increases. At a given point in
surgical setting have not been well-studied. time, there exist almost infinite possible actions within the
Human decision-making has been studied extensively in surgical field, organized in ascending levels of importance
the fields of neuroscience, psychology, and economics and complexity. Were true automation attempted, so-called
[11 13]. In vivo experiments have been able to demonstrate combinatorial explosion would result. The surgeon, pre-
the neurobiological underpinnings of decision-making pro- sented with immersive sensory data at the workstation, can
cesses during simple sensory-motor tasks [14]. Researchers rapidly combine multiple incomplete data sets in order to
have delineated complex neural circuits involved in the execute manipulator motion. This is the basis for efficacy
processes of choice valuation and decision execution; within during pre-clinical studies and success in the early clinical
associated regions, distinct populations of neurons respond use of neuroArm.
independently to each available option [15]. The result is Small-scale anatomical variability further complicates
a non-linear computational system capable not only of the process of automation by adding layers of ambiguity
Intra-operative Robotics: NeuroArm 235
and disorganization to the surgical data set, thereby increas- References

ing computational density. Though intra-operative imaging
and intuitive HMI design attempt to minimize this uncer- 1. Ballantyne GH, Moll F (2003) The da Vinci telerobotic surgical
system: the virtual operative field and telepresence surgery. Surg
tainty, the surgeon’s decision-making is not paralyzed by Clin North Am 83:1293 1304, vii
incomplete data, as some degree of sensory uncertainty is 2. Lwu S, Sutherland GR (2009) The development of robotics for
tolerated. Surgeons routinely apply knowledge of antici- interventional MRI. Neurosurg Clin N Am 20:193 206
pated structural relationships to actively guide dissection, 3. Sutherland GR, Latour I, Greer AD, Fielding T, Feil G, Newhook P
(2008) An image guided magnetic resonance compatible surgical
prior to full visualization of the relevant anatomy. Such robot. Neurosurgery 62:286 292
action is predicated on computation that allows the brain to 4. Stefanidis D, Wang F, Korndorffer JR Jr, Dunne JB, Scott DJ
both infer probability from uncertain data and determine the (2009) Robotic assistance improves intracorporeal suturing perfor
threshold of sensory certainty required to initiate action [17]. mance and safety in the operating room while decreasing operator
workload. Surg Endosc 18:18
The use of a foot-pedal to halt unintended manipulator 5. Yang T, Shadlen MN (2007) Probabilistic reasoning by neurons.
motion takes advantage of the human capacity to evaluate Nature 447:1075 1080
the error in the setting of contextual dissonance, though the 6. Ewing DR, Pigazzi A, Wang Y, Ballantyne GH (2004) Robots in
system software may not recognize technical failure. While the operating room the history. Semin Laparosc Surg 11:63 71
7. Sutherland GR, Kaibara T, Louw D, Hoult DI, Tomanek B,
neuroArm is capable of extremely basic automation in the Saunders J (1999) A mobile high field magnetic resonance system
form of a partially-automated tool exchange, it is capable of for neurosurgery. J Neurosurg 91:804 813
producing only a single solution to generate a given move- 8. Sutherland GR, Latour I, Greer AD (2008) Integrating an image
ment. Surgeons, on the other hand, can intuitively alter guided robot with intraoperative MRI: a review of the design and
construction of neuroArm. IEEE Eng Med Biol Mag 27:59 65
speed, force, and hand position efficiently and effectively 9. Pandya S, Motkoski JW, Serrano Almeida C, Greer AD, Latour I,
in order to complete a task in a variety of approaches. Sutherland GR (2009) Advancing neurosurgery with image guided
Technological developments have presented neurosur- robotics. J Neurosurg 17:17
geons with increasingly informative imaging modalities 10. Greer AD, Newhook PM, Sutherland GR (2008) Human machine
interface for Robotic surgery and stereotaxy. IEEE ASME Trans
and tools that have substantially altered the scope and nature Mechatron 13:355 361
of neurosurgical intervention. NeuroArm combines advanced 11. Dehaene S, Spelke E, Pinel P, Stanescu R, Tsivkin S (1999)
image guidance, robotic accuracy and precision, and the Sources of mathematical thinking: behavioral and brain imaging
neural processing mechanisms employed during human evidence. Science 284:970 974
12. Padoa Schioppa C, Assad JA (2006) Neurons in the orbitofrontal
control. The master-slave organization of this system links cortex encode economic value. Nature 441:223 226
human executive function and robotic accuracy in a way that 13. Platt ML, Huettel SA (2008) Risky business: the neuroeconomics
may generate surgical outcomes beyond those currently of decision making under uncertainty. Nat Neurosci 11:398 403
possible. 14. Resulaj A, Kiani R, Wolpert DM, Shadlen MN (2009) Changes of
mind in decision making. Nature 461:263 266
15. Furman M, Wang XJ (2008) Similarity effect and optimal control
Conflict of interest statement Dr. G.R. Sutherland and A.D. Greer
of multiple choice decision making. Neuron 60:1153 1168
hold shares in IMRIS (Winnipeg, Canada). M.J. Lang declares no
16. Denning PJ, Tichy WF (1990) Highly parallel computation.
conflict of inerest.
Science 250:1217 1222
17. Knill DC, Pouget A (2004) The Bayesian brain: the role of
Acknowledgements This work was supported by grants from the uncertainty in neural coding and computation. Trends Neurosci
Canada Foundation for Innovation, Western Economic Diversification, 27:712 719
and Alberta Advanced Education and Technology.
Clinical Requirements and Possible Applications of Robot Assisted
Endoscopy in Skull Base and Sinus Surgery
K.W.G. Eichhorn and F. Bootz
Abstract Functional Endoscopic Surgery of Paranasal sinu- hand. He has only his second hand free for surgical instru-
ses (FESS) and Skull Base surgery is one of the most frequent ments, suction or navigation devices. This can result in
surgeries performed at the ENT department of the Bonn unsteady endoscope images caused by tiredness or in fre-
University, Germany. Beside of surgical Navigation Robotic quent instrument changes. Since this is also true for other
is one of the upcoming fields of Computer assisted Surgery endoscopic operations like laparoscopy, robot systems are
developments. This work presents novel research and con- developed taking control of the endoscope during surgical
cepts for Robot Assisted Endoscopic Sinus Surgery (RASS) procedures (some of them are listed in [1 3]).
of the Paranasal sinuses and the anterior Skull Base contain- Our aim in Robot Assisted Endoscopic Sinus Surgery
ing the analysis of surgical workflows, the segmentation and (RASS) is that the robot guides the endoscope during an
modelling of the Paranasal sinuses and the anterior Skull operation as autonomous as possible. One of the main chal-
Base and the development of the robotic path planning. An lenges is the close proximity of critical regions like the
interdisciplinary group of software engineers and surgeons in anterior skull base, orbits, carotid arteries and optical nerves
Braunschweig and Bonn, Germany are approximate to solu- to the work space of the robot.
tions by a clinical and technical research program financed
through the DFG (Deutsche Forschungsgemeinschaft,
German research Community).
Clinical Requirements
Keywords Computer assisted surgery Endoscopic surgery
Endoscopy FESS Paranasal sinuses Robot assisted Parts of these requirements are that
surgery Robotic Skull base
1. the tip of the instrument is always in the centre of the
endoscopic view
2. the surgeon has enough free space for his instruments
3. the motion of the robot harmonize with the motion of the
Introduction surgeon
4. the robot can automatically clean the camera lens and
Over the past years Functional Endoscopic Sinus Surgery 5. the surgeon can direct the robot to (e.g. in CT data)
(FESS) has been established as one of the important standard specified locations.
techniques in the field of Otorhinolaryngology and Skull
Especially the last two points require a path planning
base Surgery. The major disadvantage of this technique is
algorithm. In contrast to most industrial applications we
that the surgeon has to hold the Endoscope by himself in one
have to deal with partly soft structures. And due to the
limited space in the nasal and paranasal cavities, the robot
has to deform some of these elastic structures like the middle
turbinates to reach the ethmoid complex. The amount of
these deformations should be restricted in such a way that
the risk of an injury is minimal for the patient. Moreover,
K.W.G. Eichhorn (*) and F. Bootz
some anatomical structures (soft or hard) are removed for
Klinik und Poliklinik fuer HNO Heilkunde/Chirurgie, Universitätsk
linikum Bonn, Sigmund Freud Strasse 25, 53105 Bonn, Germany most of the surgical approaches, which requires dynamic
e mail: klaus.eichhorn@ukb.uni bonn.de data structures.
238 K.W.G. Eichhorn and F. Bootz
Modelling the Endoscopic Sinus Surgery Results

The primer experiments were analysing regular FESS opera- As result we determine an amount of points where the
tions done by clinical specialist in a conventional two hand- endoscope tip is located during FESS which builds nearly
ed operation (One hand for the endoscope and the second the whole nasal cavity. We defined a geometrical centre as a
hand for an instrument or navigation pointer or the suction standard position for the robot to have the shortest distance
device). to reach all positions in the shortest time. Furthermore we
reduced our data to 50% of the points lying close to the
balance center and designed so a cube defining the ground
work space of the endoscope tip. The dimension of the cube
in our FESS was 16.59 mm11.38 mm6.30 mm and
Methods/Material 1.19 cm3.
As a main result we get the pose (position and orienta-
During 23 FESS procedures (five cadaver dissection and tion) of the endoscope in relation to the patients head and the
18 in vivo FESS) we mounted force/torque sensor (Nano points of passage into the nose. So we defined a plane we call
43 SI-36-0.5, Schunk GmbH &Co. KG, Lauffen, Neckar; pivot plane where all poses of the endoscope entering the
Germany) between the camera/handle and the endoscope nose are located. We also reduce these points to the closed
while operating to monitor the forces and torques occurring 50% of points to a balance center. So we get a pivot plane of
at the endoscope in six degrees of freedom. In addition 3.93 mm2.31 mm or rather 9.07 mm2 surface. The surface
we added in four FESS marker spheres at the patient, the of 9.07 mm2 is the surface of the lateral cut of all standard
endoscope and the instruments to determine the position endoscope used in sinus surgery. So in technical terms we
of these three by a self developed passive navigation sys- define a pivot point at the nasal entrance (Fig. 2).
tem and collected the endoscope images and took a cam- Our force/torque senor delivers a force/operation time
corder video from the hole procedure from a half bird view diagram for every FESS and we calculated the average
(Fig. 1). force and the maximum force and thus we found the 94.9%
Fig. 1 Screenshot of the data acquire software during conventional endoscopic sinus surgery with explanatory notes: (a) control window,
(b) force/torque data, (c) camcorder view, (d) endoscope image, (e) time line
Clinical Requirements and Possible Applications of Robot Assisted Endoscopy in Skull Base and Sinus Surgery 239
percentile lying by 7.1 N. So most of the times the forces Also we used the software ANSYS for creating a FEM-
occurring at the endoscope tip are below 7 N and in average (Finite Element Model)-Simulation.
3.3 N. In our experiments we have seen that the amounts of
data’s need for the FEM-Model are to complex for an online
calculation. So we go ahead and designed a total new model
for the path planning which will be introduced in following.
Technical Developments In our work we distinguish between the interior and the
exterior environment. The interior environment describes
To describe the soft tissue and bone properties we designed the inner structures of the patients head. So far we concen-
different experiments on human soft tissue and cadaver trated on modelling the interior environment [4]. This interi-
specimen. Elasticity and tissue properties were measured. or environment we divided into voxels according to the
CT-scans and defined for each voxel three different proper-
ties: HARD, SOFT and FREE.
HARD structures and other critical regions (e.g. brain,
eyes) can be treated as prohibitive zones for the robot.
Contacts with SOFT structures can not be avoided. But the
amount of deformation should be limited so that the risk of
injuries is minimal for the patient. In FREE Region the
maximum speed for FESS endoscope movements is allowed.
Discussion
After throwbacks in medical robotics a few years ago [5],

there are new interdisciplinary groups [6] working at the
development of new systems to relieve surgeons during
complex operations. In our work the assistance of the robot
Fig. 2 Diagram of pivot plane, pivot point and the endoscope rotations in FESS is the main topic. The surgeon would have two
and translations around the pivot point at the nasal entrance: l is the hands free to perform saver and more precise manipulations
distance between the tip and the intersection point of the endoscope
axis and the pivot plane, (a, b) is a displacement vector within the pivot
at the patient [7]. On the market there are so far only systems
plane, a is the rotation angle around the z axis (yaw), b is the rotation for telemanipulation available for example the DaVinci-
angle around the y-axis System [8], but partly autonomous working machines are
Fig. 3 Prototype for robot

assisted endoscope guidance for
surgery of the paranasal sinuses
and the anterior skull base (GUI
Graphical User Interface, Tx 40
Robotor Type Tx40)
240 K.W.G. Eichhorn and F. Bootz
still missed. In technical ways we defined a ground work Conflict of interest statement We declare that we have no conflict of
space and a fixed pivot point at the nasal entrance. We are interest.
able to reduce the movements of the robot to guide the
endoscope in the nose to one translation and two rotation
directions. By the use of limits of the work space to a cube of
1.19 cm3 and the limiting of the forces to 7 N at the endo- References
scope tip we can protect critical regions like the anterior
skull base, orbits, carotid arteries and optical nerves by
1. Dario P, Hannaford B, Menciassi A (2003) Smart surgical tools and
injuries belonging to the endoscope tip and the robot move- augmenting devices. IEEE Trans Rob Autom 19(5):782 792
ments [4, 9]. Limits for translations and rotations of the 2. Pott P (2008) Meroda, the medical robotics database. http://www.
endoscope and simplification of the voxel based interior ma.uni heidelberg.de/apps/ortho/meroda/
environment make the run time for a path planning request 3. Taylor RH, Stoianovici D (2003) Medical robotics in computer
integrated surgery. IEEE Trans Rob Autom 19(5):765 781
at about 1 s for the worst case (single threaded, E6600 CPU, 4. Tingelhoff K, Eichhorn KWG, Wagner I, Kunkel ME, Moral AI,
unoptimized code). The paths were smooth with a good Rilk M, Wahl FM, Bootz F (2008) Analysis of manual segmentation
balance between limited deformations and straight driving in paranasal CT images. Eur Arch Otorhinolaryngol 265:1061 1070
to the goal positions of the endoscope tip (Fig. 3). 5. Schulz AP, Seide K, Queitsch C, von Haugwitz A, Meiners J,
Kienast B, Tarabolsi M, Kammal M, Jürgens C (2007) Results of
total hip replacement using the Robodoc surgical assistant system:
clinical outcome and evaluation of complications for 97 proce
dures. Int J Med Robot 3(4):301 306
Perspectives 6. Federspil PA, Plinkert PK (2004) Robotic surgery in otorhino
laryngology. Otolaryngol Pol 58:237 242
7. Briner HR, Simmen D, Jones N (2005) Endoscopic sinus surgery:
In our work we developed different approaches for the advantages of the bimanual technique. Am J Rhinol 19:269 273
realization of the partly autonomous robot assisted endo- 8. Palep JH (2009) Robotic assisted minimally invasive surgery.
scope guidance in functional endoscopic sinus surgery and J Minim Access Surg 5(1):1 7
9. Wagner I, Tingelhoff K, Westphal R, Kunkel ME, Wahl FM, Bootz
the surgery of the anterior skull base. In the future we start F, Eichhorn KWG (2008) Ex vivo evaluation of force data and
new experiments with our prototype (seen in Fig. 3) and tissue elasticity for robot assisted FESS. Eur Arch Otorhinolaryn
RASS on cadaver heads. gol 265:1335 1339
Robotic Technology in Spine Surgery: Current Applications
and Future Developments
Carsten Stüer, Florian Ringel, Michael Stoffel, Andreas Reinke, Michael Behr, and Bernhard Meyer
Abstract Medical robotics incrementally appears compel- decades, important efforts have been made between academ-
ling in nowadays surgical work. The research regarding an ic and industry partnerships to develop robots suitable fur
ideal interaction between physician and computer assistance use in the operating room environment. Although some
has reached a first summit with the implementation of applications have been successful in areas of laparoscopic
commercially available robots (Intuitive Surgical’s1 da surgery at cardiac and urologic procedures, such as Intuitive
Vinci1). Moreover, neurosurgery and herein spine surgery Surgical’s da Vinci1, cerebral and spinal Neurosurgery still
- seems an ideal candidate for computer assisted surgery. presents a major challenge due to the eloquence of the
After the adoption of pure navigational support from brain surrounding anatomy. The development of minimally inva-
surgery to spine surgery a meanwhile commercially available sive surgery and introduction of medical imaging techniques
miniature robot (Mazor Surgical Technologies’ The Spine (virtual reality) is a paradigm shift for the surgical applica-
Assist1) assists in drilling thoracic and lumbar pedicle tion of medical robots for reproducibility and improved
screws. Pilot studies on efficacy, implementation into neuro- precision in surgical procedures. In many ways, spinal sur-
surgical operating room work flow proved the accuracy of gery is ideally suited for the integration of robotic-assisted
the system and we shortly outline them. Current applications surgical procedures [2]. Spinal procedures commonly
are promising, and future possible developments seem far require fine manipulation of trajectories to deeply seated,
beyond imagination. But still, medical robotics is in its critical bony structures that are accessed through a small
infancy. Many of its advantages and disadvantages must be corridor. Since spinal procedures in terms of implanting
delicately sorted out as the patients safety is of highest screws and rods constructs accompanying intervertebral
priority. Medical robots may achieve a physician’s supple- fusion can be quite lengthy and tedious, spinal surgeons
ment but not substitute. may experience fatigue, hand tremor, and a scaled down
hand motion regarding further decompressive, more
Keywords Spinal robotics Computer assisted surgery eloquent work at the spinal cord and nerve roots. Ideally,
Medical robotics Robotic neurosurgery Spine surgery robots are indefatigable, able to perform repetitious tasks
Robotic assisted screw placement with precision and reproducible outcomes.
Introduction Current Applications
Medical robotics has tremendous potential for improving the To date, medical robotics is still in its infancy and only a few
precision and capabilities of neurosurgical, and therein spi- commercial companies and their products are available.
nal surgical procedures. The intrinsic characteristics of Especially for spinal surgery, only one commercial CE- and
robots, such as high precision, repeatability and endurance FDA-approved product exists (The SpineAssist1, Mazor
make them ideal surgeon’s assistants [1]. During the last two Surgical Technologies, Caesarea, Israel), and the number
of scientific publications is scarce.
The SpineAssist1 is a bone mounted hexapod miniature
C. Stüer (*), F. Ringel, M. Stoffel, A. Reinke, M. Behr, and B. Meyer
robot (Fig. 1). It is firmly connected to the patient’s body
Department of Neurosurgery, Klinikum rechts der Isar, Technische
Universität München, Ismaningerstr. 22, 81675 München, Germany through one of three platforms: a clamped bridge attached to
e mail: carsten.stueer@lrz.tu muenchen.de the spinous process, a Hover-T bridge attached through three
242 C. Stüer et al.
Fig. 1 Picture of The

SpineAssist1 (Mazor Surgical
Technologies, Caesarea, Israel),
the only commercial robotic
system available for spine
surgery. It is a hexapod miniature
robotic device that places
preoperatively planned
trajectories for drilling screw
canal into in vivo. As shown, it
hovers above the spine via the
Hover T, which might be
attached for minimally invasive
procedures to a spinous process
and the iliac crest
pins to the patient’s bony anatomy or a bed mount device exposure was proven. Since its introduction into the field of
with only one pin connecting the Hover-T to the patient’s spinal surgery some studies have been published and besides
spine. According to the medical robot classification of Nathoo good feasability and accuracy those authors warranted pro-
et al. [3] this system, as the majority of robot-assistants, serves spective randomized controlled trials. Preliminary results of
as a shared-control system, which allows both the surgeon such a study currently done at our institution point to:
and the robot to directly control the instruments at the
1. Considerably less radiation exposure of the surgeon
same time. Prior to surgery, there are a few basic steps in
2. An improvement of the accuracy and less pedicle breach
the operation:
with spinal instrumentation
1. Preoperative planning on a reformatted CT scan of the 3. An increase of the total duration of the spinal procedure
operated area imported to the software. due to preoperative planning, intraoperative hardware
2. Calibration of the fluoroscopy c-arm. and software breakdown and therefore
3. Intra-operative fluoroscopy images of the operated field 4. An increase of annoying factors around spinal surgery
with targets mounted to the robotic platform.
4. Merging of the intra-operative fluoroscopy images with
the pre-operative CT scan (registration).
5. Execution of the pre-operative plan. Future Projects
Further information on the system can be found else-
where [4 7]. Thus, the system provides with navigation In future, as this system currently is not only in lumbar but in
and guidance, and assists via the trajectories with drilling thoracic pedicle screw insertion service The SpineAssist will
prior to screw implantation, translaminar facet screws, guide with placing cervical pedicle screws. First attempts
vertebral biopsies or kypho- and vertebroplasties. Safety with this approach have been done at several institutions in
and accuracy have been shown in clinical pilot studies [8, 9]. Europe, so far. Complex cranio-cervical, atlanto-axial and
Results of a pilot study with this system regarding clinical subaxial pathologies (i.e. rheumatoid cases, complex cervi-
feasability, safety and integration into operative workflow at cal trauma, ankylosing spondylitis) still are challenging to
our institution confirm an excellent accuracy with a devia- the spine surgeon, as the risk of neurological damage to the
tion <2 mm to the surgeons plan in 97%. 100 lumbar pedicle spinal cord or nerve roots is quoted as 0.2 1.0%, and vascu-
screws were inserted in 19 patients (average age 66.81; lar injury may occur in up to 5% with possibly devastating
range 47 86) with degenerative disease in 25 motion seg- consequences [10, 11]. Image guidance in such procedures is
ments (Figs. 1 and 2). Only one patient underwent revision supposed to offer an increased margin of safety for lateral
according to malposition of two screws (Fig. 3). After a mass, isthmic or transarticular screw placement [12]. To our
steep learning curve, easy use of the system and unconven- mind, robotic assistance in guidance-optimized placement
tional integration into the work flow was possible, as shown of cervical and lumbar total disc replacements with disc
in a previous study at another institution [9]. Furthermore, arthroplasty seems to be a further ideal area of a robotic
superiority over fluoroscopy guidance in terms of radiation supplement.
Robotic Technology in Spine Surgery: Current Applications and Future Developments 243
Fig. 2 The total average position deviation was 0.34 mm from the preoperative plan, Standard deviation 1.49; total average angular deviation
was 2.7 from the preoperative plan, Standard deviation 4.22. Average position deviation was calculated also for 2 sub groups: First 9 out of
19 procedures (solid screws) 0.5 mm; Last 10 out of 19 procedures (cannulated screws) 0.24 mm. Average angular deviation was calculated
also for 2 sub groups: First 9 out of 19 procedures (solid screws) 3.4 ; Last 10 out of 19 procedures (cannulated screws) 2.23
244 C. Stüer et al.
Fig. 3 The relative position of the screw to the pedicle was assessed and graded according to the ABC method analysis: A completely within
the pedicle; B pedicle wall breach less than 2 mm; C pedicle wall breach equal to 2 4 mm; D pedicle wall breach more than 4 mm. 90 of
100 screws were graded as A; 7 of 100 screws were graded as B; 2 of 100 screws were graded as C and one screw was graded as D
Future Prospects and Developments the safety of the application of medical robotics in the
operating room and therefore has to cope his responsibility
Generally, still some disadvantages of medical robotics have following the fundamental principle in any surgeons train-
to be counteracted. To some extent, medical robots (and ing, which is ‘‘DO NO HARM’’. The ‘‘growth’’ of the medi-
herewith especially its hardware) are bulky and space occu- cal robots from nowadays’ childhood to tomorrow’s
pying characteristics that are of major concern within the adolescence will possibly follow a hierarchical scheme or
setup and work flow in an operation room. So there is need the host of tools available to surgeons, ranging from hand-
for improvement of usability and practicability around held tools to fully powered autonomous robots. As this
bringing it into clinical application which yet is fulfilled hierarchy moves towards autonomous robots, the surgeon
with the SpineAssist. Another main disadvantage that is less and less in control, and more dependent on the
accompanies almost every medical robotic system is that it mechanical and software systems of the robot [14]. These
still requires a fixed position of the patient in relation to the technological challenges need further research areas focus-
robot base. This seems of major concern in spinal surgery, as sing system architecture, software design, mechanical
for navigation and guidance on bony structures direct con- design, user interface and imaging-compatible designs.
tact to such anatomy and avoidance of soft tissue (muscles, Thus, the development of application test beds is critical to
subcutaneous fat tissue) is warranted in at least one fixing move the field forward, and the completion of such medical
point. Thus, approaches to refine such procedures and to robotics projects require a close partnership between engi-
overcome the intra-operative movement-dependent inaccu- neers and clinicians that usually is not easy to establish.
racy have been made by developing a spine frame [13] or by Future development of medical robotics will base on
interfacing infrared systems allowing dynamic referencing integrated systems with a single interface to switch between
and complementarities [1]. Thus, in future, medical robotics planning, navigation, and robotic mode enabling the surgeon
will be increasingly linked to imaging modalities, while to have direct access to the surgical field and control of the
these systems are, for the most part, still under the direct integrated system [1]. But it is at this stage we must take into
control of the physician. This is based on the popularity and account, that the surgeon does become more of an observer
necessity of the image-guided interventions and the require- than a controller, and from a purely neurosurgical soulful
ment of the robotic systems to work within the constraints of point of view one might put this development into question.
such various modalities as CT, MRI and PET. Repeatability Nevertheless, as with the society of our countries, the
and accuracy are of major concern regarding safety issues neurosurgical field is technology-driven and technology-
facing spinal surgery. Since the robotics’ excellent accuracy dependent. Future generations of neurosurgeons will satisfy
in in vivo spinal procedures could be proven, medical robot- the immensely fast growing market for medical robotics and
ics in general will never be a supplant but always be a computer-assisted surgical equipment (average annual
supplement to the surgeon. Medical robots must operate in growing rate in the U.S. in 2004 of 32.6%) with a demand
cooperation with physicians to be fully effective [14, 15]. we can only imagine at this point. Who would have sus-
Safety measures that can be taken include the use of pected 30 years ago that we can navigate endoscopically
redundant sensors, the design of special-purpose robots within our third ventricle or a bulging disc at the segment
whose capabilities are tailored to the task at hand and the L4/5?
use of fail-safe techniques so that if the robot does fail it can
be removed and the procedure can be completed by the Conflict of interest statement We declare that we have no conflict
of interest.
surgeon. For the surgeon naturally is very concerned about
Robotic Technology in Spine Surgery: Current Applications and Future Developments 245
References 8. Lieberman IH, Togawa D, Kayanja MM, Reinhardt MK,

Friedlander A, Knoller N, Benzel EC (2006) Bone mounted mini
1. Zamorano L, Li Q, Jain S, Kaur G (2004) Robotics in neurosur ature robotic guidance for pedicle screw and translaminar facet
gery: state of the art and future technological challenges. Int J Med screw placement: part I technical development and a test case
Robot 1:7 22 result. Neurosurgery 59:641 650
2. Barzilay Y, Kaplan L, Libergall M (2008) Robotic assisted spine 9. Pechlivanis I, Kiriyanthan G, Engelhardt M, Scholz M, Lucke S,
surgery a breakthrough or a surgical toy? Int J Med Robot Harders A, Schmieder K (2009) Percutaneous placement of pedicle
4:195 196 screws in the lumbar spine using a bone mounted miniature robotic
3. Nathoo N, Cavusoglu MC, Vogelbaum MA, Barnett GH (2005) In system: first experiences and accuracy of screw placement. Spine
touch with robotics: neurosurgery for the future. Neurosurgery (Phila Pa 1976) 34:392 398
56:421 433 10. Graham JJ (1989) Complications of cervical spine surgery. A five
4. Barzilay Y, Liebergall M, Fridlander A, Knoller N (2006) Minia year report on a survey of the membership of the Cervical Spine
ture robotic guidance for spine surgery introduction of a novel Research Society by the Morbidity and Mortality Committee.
system and analysis of challenges encountered during the clinical Spine (Phila Pa 1976) 14:1046 1050
development phase at two spine centres. Int J Med Robot 11. Zeidman SM, Ducker TB, Raycroft J (1997) Trends and complica
2:146 153 tions in cervical spine surgery: 1989 1993. J Spinal Disord
5. Shoham M, Lieberman IH, Benzel EC, Togawa D, Zehavi E, 10:523 526
Zilberstein B, Roffman M, Bruskin A, Fridlander A, 12. Kelleher MO, McEvoy L, Nagaria J, Kamel M, Bolger C (2006)
Joskowicz L, Brink Danan S, Knoller N (2007) Robotic assisted Image guided transarticular atlanto axial screw fixation. Int J Med
spinal surgery from concept to clinical practice. Comput Aided Robot 2:154 160
Surg 12:105 115 13. Thomale UW, Kneissler M, Hein A, Maetzig M, Kroppenstedt SN,
6. Sukovich W, Brink Danan S, Hardenbrook M (2006) Miniature Lueth T, Woiciechowsky C (2005) A spine frame for intra
robotic guidance for pedicle screw placement in posterior spinal operative fixation to increase accuracy in spinal navigation and
fusion: early clinical experience with the SpineAssist. Int J Med robotics. Comput Aided Surg 10:151 155
Robot 2:114 122 14. Davies B (1996) A discussion of safety issues for medical robots.
7. Togawa D, Kayanja MM, Reinhardt MK, Shoham M, Balter A, In: Taylor RH, Lavallée S, Burdeaa GC, Mösges R (eds) Computer
Friedlander A, Knoller N, Benzel EC, Lieberman IH (2007) integrated surgery: technology and clinical applications. MIT Press,
Bone mounted miniature robotic guidance for pedicle screw and Cambridge, MA, pp 287 296
translaminar facet screw placement: part 2 evaluation of system 15. Elder MC, Knight JC (1995) Specifying user interfaces for safety
accuracy. Neurosurgery 60:ONS129 139 critical medical systems. Medical robotics and computer assisted
surgery. Wiley Liss, New York, pp 148 155
Microscope Integrated Indocyanine Green Video-Angiography
in Cerebrovascular Surgery
Reza Dashti, Aki Laakso, Mika Niemelä, Matti Porras, and Juha Hernesniemi
Abstract Microscope integrated indocyanine green video- microneurosurgical management of intracranial aneurysms
angiography (ICG-VA) is a new technique for intraoperative and arteriovenous malformations is well known [1 12].
assessment of blood flow that has been recently applied to Microvascular Doppler and ultrasonic perivascular flowme-
the field of Neurosurgery. ICG-VA is known as a simple and try are among other methods widely used [13 16]. Indo-
practical method of blood flow assessment with acceptable cyanine green video-angiography (ICG-VA) is a safe, and
reliability. Real time information obtained under magnifica- reliable method for assessment of blood flow which is
tion of operating microscope has many potential applica- recently introduced to the field of cerebrovascular surgery
tions in the microneurosurgical management of vascular [17].
lesions. This review is based on institutional experience This review is based on institutional experience with use
with use of ICG-VA during surgery of intracranial aneur- of ICG-VA during microneurosurgical management of cere-
ysms, AVMs and other vascular lesions at the Department of brovascular lesions. Microscope integrated ICG-VA (Opmi
Neurosurgery at Helsinki University Central Hospital. Pentero Carl Zeiss Ltd. Oberkochen, Germany) has been
in routine use for the last four years in the Department of
Keywords Arteriovenous malformations Indocyanine Neurosurgery at Helsinki University Central Hospital.
green Intracranial aneurysm Intraoperative angiography During this period ICG-VA was used during microneuro-
Surgery surgical management of more than 1,200 intracranial aneur-
ysms, 120 AVMs and some other vascular lesions.
Introduction
Indocyanine Green Video-Angiography
Intra-operative monitoring of blood flow may play an im-
The use of indocyanine green video-angiography in cerbro-
portant role during microneurosurgical management of vas-
vascular surgery was introduced by Raabe et al. in 2003 [17].
cular lesions. Various available methods of intraoperative
The technique is based on obtaining high-resolution and
blood flow measurement can provide real time information
high-contrast images detected by near infrared (NIR) camera
and improve the efficacy of treatment. Intraoperative
integrated to operating microscope. After intravenous injec-
angiography is known as the gold standard and its role during
tion of the ICG dye its florescence is induced and detected by
the NIR video camera. The result is a real time assessment of
R. Dashti cerebral vasculature under magnification of operating micro-
Department of Neurosurgery, Helsinki University Central Hospital,
00260 Helsinki, Finland scope. ICG-VA has arterial, capillary and venous phases
Department of Neurosurgery, Istanbul University, Cerrahpaşa Medical [17 19]. A dose of 0.2 0.5 mg/kg is recommended for
Faculty Hospital, Istanbul, Turkey ICG-VA, with a maximum daily dose limit of 5 mg/kg.
A. Laakso, M. Niemelä, and J. Hernesniemi (*) ICG-VA is now available and in routine use in many
Department of Neurosurgery, Helsinki University Central Hospital, centers. The technique is believed to be a safe, practical
00260 Helsinki, Finland
e mail: juha.hernesniemi@hus.fi
and cost-effective method of real time assessment of
blood flow. ICG-VA can be used during microneurosurgical
M. Porras
Department of Radiology, Helsinki University Central Hospital, management of intracranial aneurysms, AVMs, and other
00260 Helsinki, Finland vascular lesions of brain and spinal cord. Similarly, together
248 R. Dashti et al.
with microvascular Doppler and intraoperative angiography, Recently we published our experience with ICG-VA
ICG-VA can be a useful tool during surgery of vascular during microneurosurgical treatment of 239 intracranial
tumors and skull base lesions with close relation to major aneurysms in 190 patients [18]. Intraoperative ICG-VA
arteries. One of the unique advantages of the ICG-VA is assessment of total occlusion of the aneurysms and patency
its ability to assess the blood flow in perforating arteries of major or perforating arteries were retrospectively com-
[18 21]. The venous phase of the ICG-VA is another impor- pared with postoperative CTA and/or digital subtraction
tant feature which is particularly helpful in preserving the angiography (DSA). A total of 457 ICG-VA applications
veins during various steps of dissection and/or retraction were performed (1 5 for each aneurysm). Technical quality
[17, 18]. This may play the key role during surgical resection of ICG-VA was optimal for 218 aneurysms (91%) (Fig. 1).
of vascular malformations [22]. Deep location, giant size, and arachnoid scarring due to
previous operations were responsible for inadequate quality
in the rest of them.
In 14 aneurysms (6%), unexpected neck residuals were
ICG-VA During Surgery of Intracranial
detected. This rate was significantly higher in deep seated
Aneurysms aneurysms (anterior communicating or basilar artery loca-
tions). The effect of deep location of the aneurysm in the
Microneurosurgical clipping is known as a durable method surgical field was statistically significant. However, this was
of treating for intracranial aneurysms (IAs). A perfect clip not the case with the size or ruptured status of the aneurysm.
should occlude the aneurysm completely while the blood Unexpected occlusion of major branching or perforating
flow in the major and perforating branches is preserved.
Detection of neck remnant or inadvertent vessel occlusion
necessitates re-exploration if not already late. Intraoperative
detection of improperly placed clip brings the advantage of
immediate replacement to achieve complete occlusion of
or to enhance the blood flow in a compromised vessel.
Intraoperative angiography is suggested as the gold standard
and the most reliable method to control the quality of clip-
ping. However, limited routine availability, relatively high
cost and [2, 4 6, 8, 9, 12, 23, 24] a complication rate of up to
3.5% are the major limitations [12, 25, 26]. Microvascular
Doppler and ultrasonic perivascular flow probe are other
methods of blood flow assessment [13 16]. Patency of per-
forating arteries, however, cannot be detected by above
techniques.
The first application of microscope integrated ICG-VA
during aneurysm surgery was described by Raabe et al. [17].
ICG-VA was used during microneurosurgical clipping of
12 intracranial aneurysms and two patients with dural
fistulas. They reported postoperative imaging studies to be
comparable with ICG-VA findings in all patients (100%).
The technique was reported as a simple and safe alternative
to other intraoperative methods of blood flow assessment.
In their next study Raabe and colleagues [19] compared
the findings of ICG-VA with intra- or postoperative DSA
during surgical treatment of 114 patients with 124 aneur-
ysms in two neurosurgical centers. Their results revealed
90% correlation between ICG-VA and intraoperative DSA
in 60 aneurysms. Intraoperative findings of the technique
were reported to be comparable with 90% of postoperative
DSA for another 45 aneurysms [19]. de Oliveira and colleagues Fig. 1 ICG VA images of an upward projecting anterior communicat
ing artery aneurysm approached from left side. (a) Before clipping, (b)
[20] demonstrated the advantage of ICG-VA in intraopera- After successful clipping. A1 Proximal segment of anterior cerebral
tive assessment of the patency of perforating arteries around artery; A2 Post communicating segment of anterior cerebral artery;
the aneurysm. RAH Recurrent artery of Heubner, L Left, R Right
Microscope Integrated Indocyanine Green Video-Angiography in Cerebrovascular Surgery 249
arteries was found in 15 aneurysms (6%). Aneurysms locat-

ed on middle cerebral artery surprisingly constituted the
majority of them (n:10¼67%). Other locations were anterior
communicating, internal carotid-posterior communicating,
and ICA posterior wall. Location, size and ruptured status
of the aneurysm did not significantly affect the rate of
unexpected branch occlusions [18]. Usefulness of ICG-VA
was concluded in two other recent reports by Ma et al. [27]
and Li et al. [28].
ICG-VA has limited ability to visualize the part of the
base behind the aneurysm dome in deeply located aneur-
ysms. Presence of blood clots in the field or arachnoid
scarring are further restrictions. Based on our experience
intraoperative DSA and or microvascular Doppler should
be considered for verification of ICG-VA findings for deep
sited, giant, thick walled and complex aneurysms.
ICG-VA During Surgery of Brain

Arteriovenous Malformations
During microneurosurgical treatment of brain AVMs, intra-

dural strategy includes various steps of intraoperative orien-
tation, localization of the lesion, identification of the arterial
feeders, and preservation of the draining veins [29]. In
our opinion ICG-VA can serve well during early stages of
Fig. 2 ICG VA images of an AVM located at right posterior temporal
intraoperative orientation and localization of the vessels.
lobe. At the beginning of intradural part superficial veins are identified.
The technique is able to demonstrate the superficial arterial (a) Early draining vein (large arrow), (b) Late filling of cortical veins
feeders, early draining veins as well as normal ones (Fig. 2). (small arrows)
Obtained images can be compared carefully with pre-
operative angiographic studies. This helps the orientation
of surgeon under magnification of operating microscope. Recently, Killory et al., reported their experience with the
However, use of this technique during AVM surgery is use of ICG-VA during surgical resection of 10 brain AVMs
limited to that part of lesion which is already exposed and [22]. The technique was found to be useful in early detection
illuminated in the field of microscope. We find ICG-VA very of feeding arteries and veins in nine out of ten cases (90%).
helpful in case of AVMs with cisternal component such as The authors reported the ICG-VA to be not useful in detect-
paracallosal or parasylvian ones where the major feeding ing the nidus residuals or early draining veins.
arteries are in close relation to draining veins as well as the We use ICG-VA during micro-neurosurgical manage-
normal vessels. Here the early identification of large arterial ment of various other vascular lesions and as well vascular
feeder(s) and their temporary occlusion can facilitate the tumors. During surgery of spinal AVMs the technique is
later steps of dissection. Comparison of the transit time of helpful in identification of arterialized veins and fistula
the ICG dye between the arterial and venous phases can also sites. Similar findings are indicated in the reports by Colby
give an idea about the state of blood flow in the AVM during et al. [30] and Hettige et al. [31]. The usefulness of ICG-VA
surgery. ICG-VA can be repeated safely within the limits of in assessment of the patency of EC-IC bypass procedures
daily dose. We have not encountered any ICG related com- was studied by Woitzik et al. [32] and Penã-Tapia et al. [33].
plications during surgery of AVMs. In conclusion, ICG-VA can be used as a simple and
Importantly, in AVM surgery, the technique can be only practical method intraoperative blood flow assessment.
useful in early stages of superficial and sulcal dissection. We A careful interpretation of the real time information obtained
do not recommend using ICG-VA in detection of residual by ICG-VA can serve as a useful tool to improve the quality
AVMs. Intraoperative or postoperative DSA remains the of microneurosurgical management of various cerebrovas-
gold standard method in detecting residual AVMs. cular lesions.
250 R. Dashti et al.
Conflict of interest statement We declare that we have no conflict of for intraoperative assessment of vascular flow. Neurosurgery
interest. 52:132 139
18. Dashti R, Laakso A, Porras M, Niemelä M, Hernesniemi J (2009)
Microscope integrated Near infrared indocyanine green video
angiography during surgery of intracranial Aneurysms: Helsinki
References 19.
experience. Surg Neurol 71:543 550
Raabe A, Nakaji P, Beck J, Kim L, Hsu F, Kamerman JD,
Seifert V, Spetzler RF (2005) Prospective evaluation of surgical
1. Anegawa S, Hayashi T, Torigoe R, Harada K, Kihara S (1994) microscope integrated intraoperative near infrared indocyanine
Intraoperative angiography in the resection of arteriovenous mal green videoangiography during aneurysm surgery. J Neurosurg
formations. J Neurosurg 80:73 78 103:982 989
2. Barrow D, Boyer K, Joseph G (1992) Intraoperative angiography 20. de Oliveira M, Beck J, Seifert V, Teixeria M, Raabe A (2007)
in the management of neurovascular disorders. Neurosurgery Assessment of blood in perforating arteries during intraoperative
30:153 159 near infrared indocyanine green videoangiography. Neurosurgery
3. Batjer H, Frankfurt A, Purdy P, Smith S, Samson D (1988) Use of 61:ONS 63 ONS 73
etomidate, temporary arterial occlusion, and intraoperative angio 21. Raabe A, Beck J, Seifert V (2005) Technique and image quality of
graphy in surgical treatment of large and giant cerebral aneurysms. intraoperative indocyanine green angiography during aneurysm
J Neurosurg 68:234 240 surgery using surgical microscope integrated near infrared video
4. Bauer BL (1984) Intraoperative angiography in cerebral aneurysm technology. Zentralbl Neurochir 66:1 6
and AV malformation. Neurosurg Rev 7:209 217 22. Killory B, Nakaji P, Gonzales L, Ponce F, Wait S, Spetzler R
5. Chiang V, Gailloud P, Murphy K, Rigamonti D, Tamargo R (2002) (2009) Prospective evaluation of surgical microscope integrated
Routine intraoperative angiography during aneurysm surgery. J intraoperative near infrared indocyanine green angiography during
Neurosurg 96:988 992 cerebral arteriovenous malformation surgery. Neurosurgery 65:
6. Derdeyn C, Moran C, Cross D, Grubb RJ, Dacey RJ (1995) Intra 456 462
operative digital subtraction angiography: a review of 112 conse 23. Bailes J, Deeb Z, Wilson J, Jungreis C, Horton J (1992) Intrao
cutive examinations. AJNR Am J Neuroradiol 16:307 318 perative angiography and temporary balloon occlusion of the
7. Katz JM, Gologorsky Y, Tsiouris A, Wells Roth D, Mascitelli J, basilar artery as an adjunct to surgical clipping: technical note.
Gobin YP, Stieg PE, Riina HA (2006) Is routine intraoperative Neurosurgery 31:603
angiography in the surgical treatment of cerebral aneurysms justi 24. Kallmes DF, Kallmes MH (1997) Cost effectiveness of angio
fied? A consecutive series of 147 aneurysms. Neurosurgery graphy performed during surgery for ruptured intracranial aneur
58:719 727 ysms. AJNR Am J Neuroradiol 18:1453 1462
8. Klopfenstein J, Spetzler R, Kim L, Feiz Erfan I, Han P, 25. Origitano TC, Schwartz K, Anderson D, Azar Kia B, Reichman
Zabramski J, Porter RW, Albuquerque FC, McDougall CG, OH (1999) Optimal clip application and intraoperative angio
Fiorella DJ (2004) Comparison of routine and selective use of graphy for intracranial aneurysms. Surg Neurol 51:117 124
intraoperative angiography during aneurysm surgery: a prospective 26. Payner TD, Horner TG, Leipzig TJ, Scott JA, Gilmor RL, DeNardo
assessment. J Neurosurg 100:230 235 AJ (1998) Role of intraoperative angiography in the surgical treat
9. Martin N, Bentson J, Vinuela F, Hieshima G, Reicher M, Black K, ment of cerebral aneurysms. J Neurosurg 88:441 448
Dion J, Becker D (1990) Intraoperative digital subtraction angio 27. Ma C, Shi J, Wang H, Hang C, Cheng H, Wu W (2009) Intra
graphy and the surgical treatment of intracranial aneurysms and operative indocyanine green angiography in intracranial aneurysm
vascular malformations [see comment]. J Neurosurg 73:526 533 surgery: microsurgical clipping and revascularization. Clin Neurol
10. Munshi I, Macdonald RL, Weir BK (1999) Intraoperative angio Neurosurg 111:840 846
graphy of brain arteriovenous malformations. Neurosurgery 28. Li J, Lan Z, He M, You C (2009) Assessment of microscope
45:491 497 integrated indocyanine green angiography during intracranial
11. Peeters FL, Walder HA (1973) Intraoperative vertebral angiogra aneurysm surgery: a retrospective study of 120 patients. Neurol
phy in arteriovenous malformations. Neuroradiology 6:169 173 India 57:453 459
12. Tang G, Cawley C, Dion J, Barrow D (2002) Intraoperative angi 29. Yaşargil MG (1988) Microneurosurgery, vol 3 B. Georg Thieme
ography during aneurysm surgery: a prospective evaluation of Verlag, Stuttgart, New York, pp 25 53
efficacy. J Neurosurg 96:993 999 30. Colby GP, Coon AL, Sciubba DM, Bydon A, Gailloud P,
13. Amin Hanjani S, Meglio G, Gatto R, Bauer A, Charbel F (2006) Tamargo RJ (2009) Intraoperative indocyanine green angiography
The utility of intraoperative blood flow measurment during for obliteration of a spinal dural arteriovenous fistula. J Neurosurg
aneurysm surgery using an ultrasonic perivascular flow probe. Spine 11:705 709
Neurosurgery 58:ONS 305 ONS 312 31. Hettige S, Walsh D (2010) Indocyanine green video angiography
14. Bailes J, Tantuwaya L, Fukushima T, Schurman G, Davis D (1997) as an aid to surgical treatment of spinal dural arteriovenous fistulae.
Intraoperative microvascular Doppler sonography in aneurysm Acta Neurochir (Wien) 152(3):533 536
surgery. Neurosurgery 40:965 970 32. Woitzik J, Horn P, Vajkoczy P, Schmiedek P (2005) Intra
15. Charbel FT, Hoffman WE, Misra M, Ostergren LA (1998) Ultra operative control of extracranial intracranial bypass patency by
sonic perivascular flow peobe: technique and application in neuro near infrared indocyanine green videoangiography. J Neurosurg
surgery. Neurol Res 20:439 442 102:692 698
16. Firsching R, Synowitz HJ, Hanebeck J (2000) Practicability of 33. Pena Tapia PG, Kemmling A, Czabanka M, Vajkoczy P,
intraoperative microvascular Doppler sonography in aneurysm Schmiedek P (2008) Identification of the optimal cortical target
surgery. Minim Invasive Neurosurg 43:144 148 point for extracranial intracranial bypass surgery in patients with
17. Raabe A, Beck J, Gerlach R, Zimmermann M, Seifert V (2003) hemodynamic cerebrovascular insufficiency. J Neurosurg
Near infrared indocyanine green video angiography: a new method 108:655 661
Application of Intraoperative Indocyanine Green Angiography
for CNS Tumors: Results on the First 100 Cases
P. Ferroli, F. Acerbi, E. Albanese, G. Tringali, M. Broggi, A. Franzini, and G. Broggi
Abstract Purpose: To investigate the application of indo- Introduction

cyanine green (ICG) videoangiography during microsurgery
for central nervous system (CNS) tumors. Angiography with ICG has been first developed in the seven-
Methods: One hundred patients with CNS tumors who ties in Ophthalmology to evaluate choroidal microcircula-
underwent microsurgical resection from December 2006 to tion [1 6]. Recently, microscope-integrated near-infrared
December 2008 were retrospectively analyzed. The diagno- ICG videoangiography has been introduced in Neurosurgery
sis was high grade glioma in 54 cases, low grade in 17 cases, in order to visualize cerebral vessels in case of aneurysm
meningioma in 14 cases, metastasis in 12 cases and heman- clipping, bypasses and vascular malformations. It has been
gioblastoma in 3 cases. Overall, ICG was injected intrao- proven that intraoperative videoangiography with ICG may
peratively 194 times. The standard dose of 25 mg of dye help to evaluate cerebral vessels that are visible in the
was injected intravenously and intravascular fluorescence surgical field in order to get a real time diagnose of the
from within the blood vessels was imaged through an degree of aneurysm occlusion, vessel patency including
ad hoc microscope with dedicated software (Pentero, Carl the perforating arteries and, in vascular malformations, to
Zeiss Co., Oberkochen, Germany). Pre-resection and post- distinguish pathologic vessels from normal vessels and
resection arterial, capillary and venous ICG videoangio- arteries from veins based on the timing of fluorescence
graphic phases were intraoperatively observed and recorded. with the dye [7 20].
Results: ICG videangiography allowed for a good eval- To our knowledge, there are no studies in the literature
uation of blood flow in the tumoral and peritumoral exposed evaluating the potential role of ICG videoangiography dur-
vessels in all cases. No side effects due to ICG were observed. ing resection of CNS tumors. Therefore, our investigation
Conclusions: ICG video-angiography is a significant focused on whether this technique could be used as an
method for monitoring blood flow in the exposed vessels intraoperative diagnostic tool to study vascular physio-
during microsurgical removal of CNS tumors. Pre-resection pathology in CNS tumors and whether the information
videoangiography provides useful information on the tumoral retrieved could be integrated in the decision making-process
circulation and the pathology-induced alteration in surround- during surgical removal.
ing brain circulation. Post-resection examination allows for
an immediate check of patency of those vessels that are
closely related to the tumor mass and that the surgeon does
not want to damage.
Patients and Methods
Keywords Indocyanine green (ICG) videoangiography
Intracranial tumors Surgical resection Venous drainage In the period between December 2006 and December
2008, almost 1,200 patients affected by CNS tumors were
admitted at the III Neurosurgical Unit of the Department
of Neurosurgery at the Fondazione IRCCS Istituto Neurolo-
Ferroli, Acerbi and Albanese equally contributed to the paper.
gico Carlo Besta in Milan. One hundred of these patients,
P. Ferroli (*), F. Acerbi, E. Albanese, G. Tringali, M. Broggi, 67males and 33 females (mean age of 56 years), underwent
A. Franzini, and G. Broggi intraoperative ICG videoangiography and were retrospectively
Department of Neurosurgery, Fondazione Istituto Neurologico Carlo
Besta, Via Celoria 11, 20133 Milano, Italy reviewed (Table 1). All patients gave their written informed
e mail: ferrolipaolo@hotmail.com consent.
252 P. Ferroli et al.
Table 1 Types of tumors evaluated by Intraoperative ICG video surgical field and nourishing at least in part the normal
angiography in our series brain parenchyma (Fig. 1). This occurred in 3 out of 54
Tumor # Cases cases of high grade gliomas and in 4 out of 17 cases of
Meningioma 14 low-grade gliomas. Furthermore, both arterial and capillary
Low grade glioma 17
phases were useful to diagnose fronto-temporal ischemia
High grade glioma 54
Metastasis 12
after removal of a huge right frontal high grade glioma
Hemangioblastoma 3 encasing the middle cerebral artery, due to its intra-operative
thrombosis (Fig. 2).
The ICG videoangiographic pre-operative late arterial-
capillary phase provided good quality images and videos
Surgery was performed using a near-infrared videointe- of the vascular pattern of the CNS area exposed by surgery,
grated microscope (Pentero, Carl Zeiss Co., Oberkochen, well depicting both specific tumor-related alterations
Germany). The video records images from the operating and aspecific mass-effect changes. In particular, when the
microscope, illuminated with a light source including the tumor abutted the CNS surface, ICG videoangiography evi-
ICG excitation wavelength, through an optical filter that denced the pathologic characteristic of the tumor vascula-
allows only fluorescence in the ICG emission wavelength. ture, both in cases of neo-angiogenic vascular pattern and in
Only vessels directly visible in the surgical field can be cases in which the tumor showed hypoperfused or avascular
visualized. The ICG video angiography was performed areas such as in cystic or necrotic masses. In addition, in this
before and after tumor removal, following the standard pro- context, it was possible to identify artero-venous fistulas,
tocol described elsewhere [14 17, 21 29]. Briefly, ICG was which were common in high grade gliomas, and were iden-
administered intravenously by the anesthesiologist (25 mg in tified in 45 out of 54 cases (Fig. 3), typical of hemangio-
5 ml of saline). Vessel fluorescence appeared after a few blastomas (3 out of 3 cases) and sometimes observed in
seconds and was cleared within 10 min, allowing for addi- metastases (2 out of 12 cases) and rarely in meningiomas
tional injections. The resultant video was shown on the (1 out of 14 cases). Aspecific mass-effect changes i.e. brain
microscope screen in the operative room during surgery gyri compressed by the tumor and edema, and congested,
and recorded for further visualizations. were evident in a greater number of cases (all the 71 cases of
Histological diagnoses were obtained in all cases and were low and high grade tumors and 5 out of 12 cases of metas-
analyzed together with the intraoperative findings in order to tases, 5 out of 14 cases of meningiomas).
investigate the tumor-related videangiographic features. Regarding the venous phase of ICG videoangiography, this
was found useful to identify impaired venous outflow and
consequent congestion. This was evident in case of arterove-
nous fistulae, as detailed above, and in case of direct venous
compression by the tumor itself (5 out of 54 cases of high-
Results grade gliomas, 1 out of 12 cases of metastases, 3 out of 14
cases of meningiomas). In addition, ICG videoangiography
ICG Videoangiography was performed before tumor remov- allowed to identity, when present, the retrograde outflow
al in all cases. In 6 cases the operating surgeon avoided to trough anastomotic veins. These data were useful to decide
repeat post-resection ICG injection because it was consid- whether or not to cut a draining vein in order to provide a
ered useless. No adverse reaction was observed. wider and safer surgical corridor to the tumor itself (10 cases)
ICG videoangiography allowed intra-operative real- or to obtain a radical tumor resection (8 cases). Specifically, in
time assessment of the exposed vessel with excellent case of arterovenous fistulae, when normal veins afferent to
image quality and resolution. Arterial, capillary, and venous the arterialized vein did not show a retrograde flow, these
phase could be always recognized. veins were considered the only outflow for the peri-tumoral
The post-resection arterial phase was able to show, as CNS area. Therefore they were preserved in all cases in order
already demonstrated for vascular cases [7, 8, 12, 14 17, 19, to re-establish a physiologic pre-tumoral condition. On the
20], patency of big cerebral arteries that underwent manipu- contrary, when a collateral flow through an anastomotic circle
lation for the removal of tumors in contact with or fully was evident, the arterialized vein was considered un-needed
encasing them (1 planum sphenoidalis meningioma, 2 lesser and therefore was cut, without any related post-operative
wing meningiomas, and 2 sylvian metastases). In addition, complication. The same considerations were made in cases
the arterial phase was considered particularly useful to con- of a direct venous compression by the tumor.
firm patency of small arteries dissected free and preserved ICG videoangiography was also used to perform occlu-
during tumor removal because they were traversing the sion test with temporary clipping of veins that showed an
Application of Intraoperative Indocyanine Green Angiography for CNS Tumors: Results on the First 100 Cases 253
a1 a2 a3
b3
b1
b2
c1
c2 d2
Fig. 1 a1. Cystic metastasis: intraoperative view. b1. Post resection view: the middle cerebral artery and its branches (arrow) are visible under the
microscope. c1. Post resection ICG videoangiography, arterial phase: the middle cerebral artery (MCA) is well injected (arrow). a2. Pre operative
MRI (cT1): right etmoido sphenoidal meningioma with MCA encasement. b2. Dextroscope 3D reconstruction to emphasize the encasement of the
MCA. c2. Intraoperative view: post resection image showing the right internal carotid artery, the anterior and middle cerebral arteries. d2. ICG
videoangiography, arterial phase: the right internal cerebral artery (red arrow), the anterior (blue arrow) and middle (green arrow) cerebral arteries
are patent. a3. Post resection microscopic view in a case of righ rolandic high grade glioma: the artery over the resected tumor is picked up. b3.
Post resection ICG videoangiography: the artery is well injected and patent
intimate relationships with the tumors. This test was used to Discussion
evaluate the presence of anastomotic circle allowing for
venous sacrifice in order to increase the degree of tumor The intraoperative ICG videoangiography is a relatively
resection. new technique of intraoperative investigation that has been
A summary of videoangiographic results in different recently applied in the field of Vascular Neurosurgery.
tumors is shown in Table 2. Routine or selective use of this technique during surgery of
Fig. 2 (a) Pre operative MRI (cT1) showing a fronto temporal GBM on the right side in close relation with the homolateral middle cerebral artery
(MCA). (b) Intraoperative view after right fronto temporal approach and dural opening. Although mannitol was given to the patient, under the
microscope the exposed cortex appears congested due to the edema. (c) Pre resection ICG videoangiography: a stagnant flow within the cortical
vessels is detected (arrows). (d) Post resection view. After tumor removal the surrounding brain appears decompressed. (e) Post resection ICG
videoangiography showing no injection of the exposed area of the frontal lobe. An ICP monitoring was therefore positioned (f) Due to the
increasing of the ICP the patient underwent an early post operative CT scan. A huge ipodensity of the right hemisphere associated with shift of the
structure of the midline was detected. An hemicraniectomy was therefore performed. (g) Post operative CT after decompressive craniectomy
Fig. 3 (a) Pre resection ICG videoangiography in one case of right frontal metastasis (intraoperative microscopic picture in the left corner): in late
arterial/capillary phase early injection of arterialized venous vessels (red arrows), artero venous (A V) fistula, together with physiologic injection
of the capillars are visible. The dye is not yet filling the non arterialized vein (blue arrow). (b) Pre resection ICG videoangiography in one case of
left frontal high grade glioma (intraoperative microscopic picture in the left corner): in late arterial/capillary phase an A V fistula is present (red
arrow)
vascular malformations and revascularization procedures in order to get a real time diagnose of the degree of aneurysm
has been suggested by many authors. It has been proven occlusion and vessel patency including the perforating
that intraoperative videoangiography with ICG may help to arteries. In cases of vascular malformations, it can be used
evaluate cerebral vessels that are visible in the surgical field to distinguish pathologic vessels from normal vessels and
Table 2 ICG videoangiographic characteristics identified in the three phases, specified for each tumor
Tumor Arterial phase Capillary phase Venous phase
Meningioma Post resection vessel patency (3/14) Pre resection A V shunt (1/14) Pre resection A V shunt (1/14)
Low grade glioma Post resection vessel patency (4/17)
High grade glioma Post resection vessel patency (3/54) Pre resection A V shunt (45/54) Pre resection A V shunt (45/54)
Post resection ischemia (1/54)
Metastasis Post resection vessel patency (2/12) Pre resection A V shunt (2/12) Pre resection A V shunt (2/12)
Hemangioblastoma Pre resection A V shunt (3/3) Pre resection A V shunt (3/3)
arteries from veins based on the timing of fluorescence with capillary, and venous phase could be recognized in all cases.
the dye [7 20]. Therefore, the neovascular architecture, alteration of the
To our knowledge, there are no studies in the literature calibre, morphology and course of vessels, and the haemo-
investigating the potential role of ICG videoangiography dynamic patterns could be studied. The ICG transit time in
during resection of CNS tumors. The angiographic evalua- tumoral vessels was found to be normal or shortened such as
tion of CNS tumor vasculature before surgical removal was in case of malignant tumor (i.e. high grade gliomas and
the only pre-operative radiological information before the metastases). A short flow-time, due to pathological low-
advent of Computerized Tomography and, more recently, resistance vessels that results in arteriovenous shunting,
Magnetic Resonance. In the early 1920s, Egas Moniz intro- was found to be common in high grade gliomas, as is the
duced and developed the idea of utilizing X-rays as a method presence of neovascular architecture, dysplastic vessels and
of making visible the blood vessels of the brain and to locate thrombosed veins, as previously demonstrated with tradi-
brain tumors. After mapping the normal distribution of the tional cerebral angiography [21, 37, 38]. Although the arte-
intracranial blood vessels, he clinically used his method riovenous shunts were not always found (45 out of 54 high
in 1927, outlining with X rays the location and size of grade gliomas), an early visualization of the venous com-
a patient’s brain tumor by the tumor’s displacement of partment was always present. Because arteriovenous shunt-
injected arteries[25 27]. At that time he defined the angio- ing causes an arterial steal that leads to hypoperfusion of the
graphic phases that are still used today: arterial, capillary, surrounding parenchyma, it has been speculated that this
early and late venous phase. Direct and indirect signs were mechanism could contribute to local ischemia, eventually
used to diagnose a brain tumor. The direct signs, i.e. mor- causing tissue necrosis. The presence of multiple areas of
phological features of the pathological vessels, were utilized necrosis suggests that the neovasculature fails to supply the
to define the nature of the lesion and the indirect signs, i.e. rapidly growing tumor tissue [24]. Sometimes, superficial
the displacement of the cerebral arteries due to the lesion, avascular areas in case of high grade glioma and metastasis,
were used to locate the tumor [22, 28 31]. Nowadays, the have been seen during pre-resection ICG videoangiography.
diagnosis of CNS tumors is based mainly on computed Although the presence of these vascular characteristics
tomography and magnetic resonance imaging, however, of GBMs have been known for decades, in the last WHO
selective cerebral angiography is still utilized when the classification there are still controversies on arteriovenous
neurosurgeon is considering a combined endovascular- shunts and particularly on the presence/absence of necrosis
surgicalstrategy of treatment in selected cases and it’s still in GBMs [23, 39]. Although the number of cases studied is
the gold standard to evaluate the patency of the dural sinus in low, all the hemangioblastomas, which are peculiar types of
oncological cases [32 36]. tumor, showed the presence of superficial pathologic vessel
In this study we had the opportunity to evaluate whether architecture, with arterovenous fistulae, and arterialized and
ICG videoangiography could provide some information dilated vein draining the tumor nodule.
regarding vascular physiopathology of CNS tumors. We Despite the permeability of the blood brain barrier, the
studied a heterogeneous population consisting mainly on dye does not penetrate the membrane and we were unable to
high grade gliomas, but including also low grade gliomas, define the margins of the tumor in gliomas and metastases.
meningiomas, metastases and hemangioblastomas. As Another opportunity offered by this study was to investi-
already known from previous studies on vascular cases, gate whether the information retrieved by the use of intrao-
ICG videoangiography allowed intra-operative real time perative ICG videoangiography could be integrated in the
assessment of the exposed vessel with excellent image qual- decision making process during surgical removal.
ity and resolution. Therefore, we could evaluate only the First of all, apart from direct visual observation and
area exposed by the craniotomy, which, especially for mini- microvascular Doppler, intraoperative ICG videoangio-
mally invasive approaches that we routinely use for most of graphy was found to be a useful tool for intraoperative
our cases, not always represented the entire area with direct assessment of post-resection vessel patency. Microscope-
and indirect signs of the tumor’s presence. However, arterial, integrated near-infrared ICG videoangiography can confirm
the presence of good flow through the arteries that have been 6. Hope Ross M, Yannuzzi LA, Gragoudas ES, Guyer DR, Slakter JS,
exposed during tumor resection. In addition, in the case of Sorenson JA, Krupsky S, Orlock DA, Puliafito CA (1994) Adverse
reactions due to indocyanine green. Ophthalmology 101:529 533
intra-operative diagnosis of stroke, as happened for one of 7. Dashti R, Laakso A, Niemelä M, Porras M, Hernesniemi J (2009)
the patient in this series (Fig. 2), the early evidence of this Microscope integrated near infrared indocyanine green videoan
complication leaded to ICU admission with ICP monitoring giography during surgery of intracranial aneurysms: the Helsinki
that allowed for further diagnosis and immediate decom- experience. Surg Neurol 71:543 550
8. de Oliveira M, Beck J, Seifert V et al (2007) Assessment of flow in
pressive craniectomy to reduce intracranial hypertension. perforating arteries during intracranial aneurysm surgery using
Particularly interesting were the pieces of information intraoperative near infrared indocyanine green videoangiography.
provided by ICG videoangiography during the venous Neurosurgery 61:ONS 63 ONS 73
phase. The possibility of post-operative complications in 9. Ferroli P, Acerbi F, Broggi M, Broggi G (2010) Arterovenous
micro malformation of the trigeminal root: intraoperative diagno
case of venous sacrifice during neurosurgical procedures sis with ICG videoangiography. Neurosurgery (in press)
has always been debated [40, 41]. However, there are not 10. Ferroli P, Tringali G, Albanese E, Broggi G (2008) Developmental
definite data on post-operative complications rate, related to venous anomaly of petrous veins: intraoperative findings and indo
the sacrifice of single veins in every patient. Although this cyanine green video angiographic study. Neurosurgery 62(5 suppl
2):ONS418 ONS421
risk can be related to the variability of individual pattern of 11. Hettige S, Walsh D (2010) Indocyanine green video angiography
venous drainage, which could be evaluated by pre-operative as an aid to surgical treatment of spinal dural arteriovenous fistulae.
angiography, it is not possible to investigate the real entity of Acta Neurochir (Wien) 152(3):533 536
anastomotic circles because an occlusive test on single veins 12. Imizu S, Kato Y, Sangli A, Oguri D, Sano H (2008) Assessment of
incomplete clipping of aneurysms intraoperatively by a near infra
can not be performed. Our data on ICG videoangiography red indocyanine green video angiography (Niicg Va) integrated
provide some insights into this discussion. In fact, we could microscope. Minim Invasive Neurosurg 51:199 203
evaluate the pattern of venous drainage of the tumor and the 13. Killory BD, Nakaji P, Gonzales LF, Ponce FA, Wait SD,
surrounding brain parenchyma in every patient. When vein Spetzler RF (2009) Prospective evaluation of surgical micro
scope integrated intraoperative near infrared indocyanine green
outflow was impaired, as in case of arterovenous fistulae or angiography during cerebral arteriovenous malformation surgery.
direct vein compression by the tumor, ICG videoangiogra- Neurosurgery 65(3):456 462
phy could offer a unique possibility to intraoperatively eval- 14. Li J, Lan Z, He M, You C (2009) Assessment of microscope
uate the presence of anastomotic circulation. This result integrated indocyanine green angiography during intracranial
aneurysm surgery: a retrospective study of 120 patients. Neurol
could be achieved through a temporary clipping test of India 57(4):453 459
veins closely related to the tumor that evaluated the ICG 15. Raabe A, Beck J, Gerlach R et al (2003) Near infrared indocyanine
clearance time of the tested vein. We used this information green video angiography: a new method for intraoperative assess
to evaluate whether vein sacrifice could safely be performed. ment of vascular flow. Neurosurgery 52:132 139
16. Raabe A, Beck J, Seifert V (2005) Technique and image quality of
Interestingly, even if these data have been retrieved on few intraoperative indocyanine green angiography during aneurysm
patients, the favourable outcome in these cases is encourag- surgery using surgical microscope integrated near infrared video
ing and the ICG videoangiography seems to add useful technology. Zentralbl Neurochir 66:1 6
information regarding individual venous variability. 17. Raabe A, Nakaji P, Beck J et al (2005) Prospective evaluation of
surgical microscope integrated intraoperative near infrared indo
cyanine green videoangiography during aneurysm surgery. J Neu
Conflict of interest statement We declare that we have no conflict of
rosurg 103:982 989
interest.
18. Takagi Y, Kikuta K, Nozaki K, Sawamura K, Hashimoto N (2007)
Detection of a residual nidus by surgical microscope integrated
intraoperative near infrared indocyanine green videoangiography
in a child with a cerebral arteriovenous malformation. J Neurosurg
References 107(5 suppl):416 418
19. Woitzik J, Horn P, Vajkoczy P et al (2005) Intraoperative control
of extracranial intracranial bypass patency by near infrared indo
1. Benson RC, Kues HA (1978) Fluorescence properties of indocya
cyanine green videoangiography. J Neurosurg 102:692 698
nine green as related to angiography. Phys Med Biol 23:159 163
20. Xu BN, Sun ZH, Romani R, Jiang JL, Wu C, Zhou DB, Yu XG,
2. Cochran ST, Bomyea K, Sayre JW (2001) Trends in adverse events
Hernesniemi J, Li BM (2009) Microsurgical management of large
after IV administration of contrast media. AJR Am J Roentgenol
and giant paraclinoid aneurysms. Surg Neurol Oct 13 [Epub ahead
176:1385 1388
of print]
3. Flower RW, Hochheimer BF (1976) Indocyanine green dye fluo
21. Kleihues P, Cavenee WK (eds) (2000) World Health Organi
rescence and infrared absorption choroidal angiography performed
zation classification of tumours. Pathology and genetics of tumours
simultaneously with fluorescein angiography. Johns Hopkins Med
of the nervous system, 2nd edn. IARC Press, Lyon, ISBN 92 83
J 138:3 42
22409 4
4. Fox IS, Wood EH (1960) Indocyanine green: physical and patho
22. Krayenbuhl H, Yasargil MG (1967) Diagnosi strutturale dei tumori
logical properties. Proc Mayo Clinic 35:732
intracranici. In: Krayenbuhl H, Yasargil MG (eds) L’angiografia
5. Hochheimer BF (1971) Angiography of the retina with indocya
cerebrale, 1st edn. Piccin, Padova, pp 263 314
nine green. Arch Ophthalmol 86:564 565
23. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, Burger PC, 33. Cushing H (1922) The meningiomas (dural endotheliomas): their
Jouvet A et al (2007) The 2007 WHO classification of tumours of source and favore seats of origin. Brain 45:282
the central nervous system. Acta Neuropathol 114:97 109 34. Cushing H, Eisenhardt L (1938) Meningiomas: their classification,
24. Mariani L, Schroth G, Wielepp JP, Haldemann A, Seiler RW regional behaviour, life history and surgical end results. Charles C
(2001) Intratumoral arteriovenous shunting in malignant gliomas. Thomas, Springfield, IL
Neurosurgery 48(2):353 357 35. Standard SC, Ahuja A, Livingston K et al (1994) Endovascular
25. Moniz E (1927) L’encéphalograpie artérielle, son imporance dans embolization and surgical excision for the treatment of cerebellar
la localisation des tumeurs cérébrales. Rev Neurol 34:72 89 and brain stem Hemangioblastomas. Surg Neurol 41:405 410
26. Moniz E (1931) Diagnostic des tumeurs cérébrals et épreuve de 36. Tampieri D, Leblanc R, TerBrugge K (1993) Preoperative emboli
l’encephalographie artérielle. Masson & Cie, Paris zation of brain and spinal hemangioblastomas. Neurosurgery
27. Moniz E (1934) L’angiographie cérébrale. Masson & Cie, Paris 33:502 505
28. Newton TH, Potts DG (eds) (1971) Radiology of the skull and 37. Feiring EH, Shapiro JH, Lubetsky HW (1963) The ring like vascu
brain, vol 2. C.V. Mosby, St. Louis lar pattern in cerebral angiography. Am J Roentgenol 89:385 390
29. Panzini A, Conte P (1983) Diagnostica angiografica dei tumori 38. Tonnis W, Walter W (1959) Das glioblastoma multiforme. Bericht
cerebrali sopratentoriali. Piccin, Padova über 2611 Fälle. Acta Neurochir Suppl 6:40 62
30. Gallingioni F (1980) Semeiologia angiografica nella patologia 39. Scheithauer BW, Fuller GN, VandenBerg SR (2008) The 2007
vascolare in Neuroradiologia Piccin, Padova, pp 398 401 WHO classification of tumors of the nervous system: controversies
31. Zachrisson L (1963) Angiography of cerebral metastases. Acta in surgical neuropathology. Brain Pathol 18:307 316
Radiol 1:521 527 40. Auque J (1996) Neurochirurgie 42(suppl 1):88 108
32. Al Mefty O (1998) Operative atlas of meningiomas. Lippincott 41. Jaeger R (1951) Observation on resection of the superior longi
Raven, Philadelphia tudinal sinus at and posterior to the rolandic and venous inflow.
J Neurosurg 8:103 109
A Technical Description of the Brain Tumor Window Model: An In Vivo
Model for the Evaluation of Intraoperative Contrast Agents
Daniel A. Orringer, Thomas Chen, Dah-Luen Huang, Martin Philbert, Raoul Kopelman, and Oren Sagher
Abstract The evaluation of candidate optical contrast Increasingly robust evidence suggests that the extent of
agents for brain tumor delineation in ex vivo models may surgical resection correlates with patient outcome for all
not accurately predict their activity in vivo. This study types of brain tumors [1]. Stereotactic navigation, intra-
describes an in vivo model system designed to assess optical operative ultrasound and intraoperative MRI have been
contrast agents for brain tumor delineation. The brain tumor developed been developed to improve the extent of resec-
window (BTW) model was created by performing biparietal tion. However, these technologies are limited by generating
craniectomies on 8-week-old Sprague-Dawley rats, injecting data that physically separated from the operative field,
9L glioma cells into the cortex and bonding a cover slip to requiring the surgeon to correlate an image with the reality
the cranial defect with cyanoacrylate glue. Tumor growth of the appearance of the operative field. To bridge the
was followed serially and occurred in an exponential fash- gap between diagnostic images and the operating field,
ion. Once tumors on the cortical surface achieved a 1 mm investigators have long proposed the use of dyes to optically
radius, intravenous contrast agents were injected while delineate tumor margins [2 9].
the appearance of the cortical surface was recorded. Experimental evaluation of tumor-delineating dyes has
Computerized image analysis was used to quantitatively been carried out exclusively in ex vivo models. However,
evaluate visible differences between tumor and normal due to the visual differences between perfused and non-
brain. Tumor margins became readily apparent following perfused tissue, we suggest that the properties of candidate
contrast administration in the BTW model. Based on red optical contrast agents could be best characterized using
component intensity, tumor delineation improved fourfold in vivo, rather than ex vivo models. Therefore, we aimed
at 50 min post-contrast administration in the BTW model to create an animal model to allow dynamic, in vivo visuali-
(P<0.002). In summary, window placement overlying an zation of the tumor brain interface. We describe a combina-
implanted glioma is technically possible and well tolerated tion of the conventional 9L implanted glioma model with the
in the rat. The BTW model is a valid system for assessing the chronic closed cranial window model to create the brain
in vivo activity of optical contrast agents. tumor window (BTW) model, a new system for evaluating
the visual appearance of experimental brain tumors in vivo.
Keywords Brain tumor Cranial window Intraoperative
imaging

D.A. Orringer, D. L. Huang, and O. Sagher (*)
Department of Neurosurgery, University of Michigan Health System, The 9L gliosarcoma cell line was cultured under standard
1500 E. Medical Center Drive, Ann Arbor, MI 48109 5338, USA cell culture conditions in RPMI media with 10% fetal bovine
e mail: osagher@umich.edu serum (InVitrogen, Carlsbad CA).
T. Chen Approval from the University Committee on Use and
University of Michigan, Ann Arbor, MI USA Care of Animals at the University of Michigan was obtained
M. Philbert prior to all experiments. Sprague-Dawley rats weighing
Department of Toxicology, University of Michigan, Ann Arbor, MI
USA
250 350 g were premedicated with buprenorphine 0.1 mg/
kg and anesthetized with inhaled isoflurane delivered
R. Kopelman
Department of Chemistry, University of Michigan, Ann Arbor, MI through a 16 gauge endotracheal tube. Vancomycin (25 mg/
USA kg) was administered intraperitoneally prior to incision.
260 D.A. Orringer et al.
After fixation in a stereotactic head holder, a midline The caudal-most dura was then incised in a radial fashion
incision was carried out over the skull from the frontal along the base of the cranial defect and reflected cranially
region to the occipito-cervical junction. Sutures were so that it could be removed in a single flap. Care was
placed 2 mm from the apex of the superior and inferior taken to avoid disrupting the large cortical veins entering
apices of the incision and placed on tension to retract the the superior sagittal sinus in the midline which are con-
scalp (Fig. 1a). The periosteum was detached from the tinuous with the dura that must be removed medially. A
skull through blunt dissection and removed using bipolar dissecting forcep with curved tines was helpful in dissect-
electrocautery. The periosteum, in continuity with fascial ing the dura without causing damage to the underlying
layers overlying the muscles of mastication, was removed structures. Once the dura was removed (Fig. 1b), the brain
to expose a sagitally-oriented ridge of bone. The bregma was irrigated with sterile normal saline. 105 9L cells were
and lambda were identified as landmarks for placement of injected at a depth of 1mm into the right or left frontal
a craniectomy (Fig. 1a). Bleeding during exposure was portion of the exposed brain in an area lacking large
controlled with bone wax, bipolar electrocautery and re- blood vessels (Fig. 1c). A thin round glass microscope
traction. A high-speed drill was used to perform a nearly cover slip (10 mm diameter) was placed over the cranial
full thickness craniectomy of the medial aspect of the defect (Fig. 1d). The edges of the cover slip were covered
parietal bones bilaterally leaving an egg-shell thickness with two layers of cyanoacrylate glue (Fig. 1d) and a
of bone overlying the dura (Fig. 1a). Minimal drilling custom cut plastic ring was sown to the scalp with 3
was carried out of the bone overlying the superior sagittal 0 nylon suture to enable continuous in vivo visualization
sinus. The residual layer of bone was removed with a fine of the implanted window (data not shown).
forceps and curved Penfield dissector. A linear rent in the Animals were weighed and monitored daily. The surface
dura in the coronal plane, towards the caudal end of the of the cortex was inspected for signs of tumor growth in the
craniectomy was created using a bent 30-gauge needle. region of the injection. Signs of tumor growth observed
Fig. 1 Essential stages in the creation of a brain tumor window model: After reflecting the scalp edges laterally, a nearly full thickness
craniectomy is performed, sparing the bone over the superior sagittal sinus (a). Once the remaining bone overlying the dura has been removed,
a small durotomy at the base of the craniectomy is created to create a larger linear dural rent which is used to peel the dura forward, ideally in a
single flap. Once the dural flap has been removed, the shiny arachnoidal surface of the brain is appreciated (b). Under microscopic guidance, a
10 mL, 26 gauge Hamilton syringe mounted to a steretactic injector is lowered through the meninges, 1 mm deep into an avasacular region of cortex
(c). After injection, the surface of the brain is irrigated to remove residual tumor cells. A 10 mm glass cover slip is carefully placed on the surface of
the brain and a confluent circumferential layer of cyanoacrylate glue, overlapping the edge of the coverslip and the craniectomy margin is applied
(d). Glue is usually dry within 45 min at which time a plastic ring stenting the scalp open for continuous observation is applied
A Technical Description of the Brain Tumor Window Model: An In Vivo Model for the Evaluation 261
include hypervascularity, pinkish color change in the region Results

of tumor injection and elevated cell mass at the site of
injection. The size of the cortical region showing signs Technique Development
of tumor growth was measured serially to generate a
tumor growth curve. Animals were imaged by MRI using The BTW model was refined using 50 animals. 49/50 ani-
previously published protocols to confirm the presence of mals demonstrated a 2 3 day period of approximately 7%
tumors [10]. body weight loss, but did not show any behavioral abnorm-
When the radius of tumors reached 1 mm, and the alities. Among the complications encountered in the initial
tumors were clearly visible adjacent to normal brain tissue, model development were, failure to thrive (resulting in death
they were used for evaluation of optical contrast agents. on the 2nd post-operative day), infection underlying the
A cut-down was performed to establish femoral venous window (8/50), no tumor growth (6/50), diffuse tumor
access. PE50 surgical tubing was placed (0.2500 inner growth (3/50) and extensive hemorrhage under the window
diameter, Dow Corning, Midland, Michigan) into the right (3/50). One animal failed to thrive, possibly due to stroke.
femoral vein. Animals were then placed into a stereotactic All other animals returned to their preoperative weight,
head holder and continuous video recording with a high- baseline neurological status and to the expected rate of
definition camcorder under visible was then initiated. growth for rats of their age by the 5th post-operative day.
Coomassie blue (CB), an optical contrast agent was Tumor growth was evaluated serially whenever possible.
administered intravenously over 5 min using a Medfusion The progression of tumor growth observed is demonstrated
3500 syringe pump (Medex, Dublin, OH) infusion syringe in Fig. 2. Assuming a spherical growth pattern, tumor
pump. growth followed an exponential curve (data not shown).
Video collected during the experiment was examined Roughly spherical tumor geometry was confirmed by MRI
qualitatively to evaluate the difference in appearance (data not shown). On average, tumor became visible after
between normal cortex and tumor tissue. Still images were four days and achieved a radius of 1 mm by 12.8 days.
generated from the video and analyzed colorimetrically with Tumor tissue appeared slightly redder than the normal corti-
Image J as previously described to quantify the degree of cal surface and has vascular markings that are distinct from
color change in the tumor [11]. The difference in red hue was normal cortex.
found to be the best method to reflect the visual difference
between tumor and normal brain.
Brain tissue was examined ex vivo following each experi-
ment on both a macroscopic and microscopic level. At the
conclusion of each experiment, the brain was removed and BTW Model Characterization
photographed digitally on a standard white background. Gross
coronal sections were photographed digitally then immersed After the BTW model could be reliably and consistently
in 4% paraformaldehyde for 15 min followed by 40% sucrose created, 6 BTW animals possessing optimal imaging char-
solution for 24 h. The sections were frozen in Tissue-Tek acteristics were used to evaluate the ability of contrast agents
O.C.T. Compound (Sakura Finetek, The Netherlands). to delineate neoplastic tissue in the BTW model. Prior to
Routine hematoxylin and eosin staining was performed on contrast administration, it was difficult to visibly delineate
15 mm coronal sections created with a cryotome. the margins or implanted tumor in comparison to adjacent
Fig. 2 Progression of tumor

growth in the brain tumor window
model. Tumor expands radially
and can be tracked following
implantation. A representative
brain tumor window model is
shown 1, 5, 9, and 13 days post
implantation
262 D.A. Orringer et al.
Fig. 3 Images of a brain tumor

window model 12 days post
implantation before (a) and
50 min after Coomassie blue
administration (b). Note the
relatively subtle difference in the
color of the tumor prior to
contrast administration and the
clear difference between the
appearance of tumor and normal
brain post contrast
normal cortex (Fig. 3). Following contrast administration, 5-aminofluorescein labeled albumin [12] have been demon-
there was a marked improvement in our ability to distinguish strated to be useful in delineating glioma tissue intra-
tumor from normal brain (Fig. 3). The calculated difference operatively with fluorescent microscopy. We have recently
in red hue, reflecting the degree of color difference between suggested that dye-loaded nanoparticles may hold promise
tumor and normal brain, improved from 47.3 pre-contrast as visible contrast agents for brain tumor delineation [13].
to 141.6 post-contrast (P<0.0002). Maximal delineation We felt that the best way to evaluate candidate contrast
was achieved within 50 min of contrast administration and agents was through a model system that would recreate the
persisted throughout the entire 2-h experiment. common difficulty surgeons face in distinguishing glioma
Whole brains harvested from animals following experi- from normal brain. We combined a common model for
ments were evaluated histologically. Grossly, there was a studying gliomas (9L gliosarcoma) with one primarily used
spherical superficial tissue mass at the site of the injection. to study the cerebral vasculature (cranial window model) to
Microscopic examination of specimens confirmed evidence create the brain tumor window model. The initial complica-
of a spherical implanted glioma adjacent to normal cortex. tions encountered in the development of the model (superfi-
Glioma tissue appeared hypercellular in comparison to nor- cial cortical hemorrhage, infection, failure or tumor growth,
mal brain. The nuclear to cytoplasmic ratio in the neoplastic excessive tumor growth and stroke) subsided with experi-
appearing tissue was much greater than that of cells identi- ence. The result of our efforts was a model that allows direct
fied in the normal cortical tissue. The nuclei of the implanted observation of a proliferating tumor in vivo with subtly
tumor appeared highly polymorphic. Mitotic figures were visible margins. From a qualitative and quantitative perspec-
easily identified. Fronds of tumor cells could be observed tive tumor margins in the BTW model are dramatically
infiltrating into normal cortex adjacent to tumor (data not clearer following the administration of contrast agents.
shown). Our data suggests that conventional ex vivo 9L glioma
Delineation of tumor was compared in the BTW and the models may overestimate the degree of delineation afforded
conventional implanted 9L glioma model (data not shown). by candidate contrast agents. We feel this observation arises
The conventional ex vivo model overestimated the degree of from the lack of perfusion in ex vivo specimens. Both
delineation by CB (P<0.03). There was a sharp change in models have tumor margins that are well delineated by
red hue at the visually apparent and MRI-defined tumor contrast and correspond to MRI-defined tumor margins.
margin in both the BTW and conventional models. Since human tumors margins are rarely as regular as those
of implanted gliomas, the utility of the BTW model might be
improved if it were performed with a more infiltrative type
of tumor.
Discussion In addition, there are a number of potential applications
of the brain tumor window model that may not relate to
Achieving gross total resection is a key component of studying visible contrast agents. First, the BTW model
current treatment of gliomas [1]. Visible contrast agents might be used for evaluating fluorescent and near-IR agents
such as fluorescein, [7] indocyanine-green [2] and bromo- for tumor delineation. Since the BTW allows direct observa-
phenol blue [6] hold promise in assisting in the delineation tion of tumor tissue in situ, it is possible that it could be
of neoplastic tissue within the operative field [3]. Fluores- used to track the response of tumor to anti-cancer therapies.
cent contrast agents such as, 5-aminolevulinic acid [8] and In summary, while we have demonstrated the utility of the
A Technical Description of the Brain Tumor Window Model: An In Vivo Model for the Evaluation 263
BTW model for evaluating visible contrast agents, it is possi- 6. Ozawa T, Britz GW, Kinder DH, Spence AM, VandenBerg S,
ble that the BTW model may assist be useful in a variety of Lamborn KR, Deen DF, Berger MS (2005) Bromophenol
blue staining of tumors in a rat glioma model. Neurosurgery
in vivo glioma studies. 57:1041 1047, discussion 1041 1047
7. Shinoda J, Yano H, Yoshimura S, Okumura A, Kaku Y, Iwama T,
Disclosures This work was supported by grants from the National Sakai N (2003) Fluorescence guided resection of glioblastoma
Institute of Biomedical Imaging and Bioengineering (1R01EB007977 multiforme by using high dose fluorescein sodium. Technical
01, to RK), the National Cancer Institute (1R21CA125297 01A1, to note. J Neurosurg 99:597 603
RK, and 1F32CA126295 01A1, to DAO), and the 2007 CNS Basic/ 8. Stummer W, Pichlmeier U, Meinel T, Wiestler OD, Zanella F,
Translational Resident Research Fellowship (to DAO). The authors Reulen HJ (2006) Fluorescence guided surgery with 5 aminolevu
have no personal financial or institutional interest in any of the drugs, linic acid for resection of malignant glioma: a randomised con
materials, or devices described in this article. trolled multicentre phase III trial. Lancet Oncol 7:392 401
9. Veiseh M, Gabikian P, Bahrami SB, Veiseh O, Zhang M,
Conflict of interest statement We declare that we have no conflict Hackman RC, Ravanpay AC, Stroud MR, Kusuma Y, Hansen SJ,
of interest. Kwok D, Munoz NM, Sze RW, Grady WM, Greenberg NM,
Ellenbogen RG, Olson JM (2007) Tumor paint: a chlorotoxin:
Cy5.5 bioconjugate for intraoperative visualization of cancer
foci. Cancer Res 67:6882 6888
10. Reddy GR, Bhojani MS, McConville P, Moody J, Moffat BA,
References Hall DE, Kim G, Koo YE, Woolliscroft MJ, Sugai JV,
Johnson TD, Philbert MA, Kopelman R, Rehemtulla A, Ross BD
(2006) Vascular targeted nanoparticles for imaging and treatment
1. Sanai N, Berger MS (2008) Glioma extent of resection and its of brain tumors. Clin Cancer Res 12:6677 6686
impact on patient outcome. Neurosurgery 62:753 764, discussion 11. Orringer DA, Koo YE, Chen T, Kim G, Hah HJ, Xu H, Wang S,
264 266 Keep R, Philbert MA, Kopelman R, Sagher O (2009) In vitro
2. Britz GW, Ghatan S, Spence AM, Berger MS (2002) Intracarotid characterization of a targeted, dye loaded nanodevice for intrao
RMP 7 enhanced indocyanine green staining of tumors in a rat perative tumor delineation. Neurosurgery 64:965 971, discussion
glioma model. J Neurooncol 56:227 232 971 972
3. Hansen DA, Spence AM, Carski T, Berger MS (1993) Indocyanine 12. Kremer P, Mahmoudreza F, Ding R, Pritsch M, Zoubaa S, Frei E
green (ICG) staining and demarcation of tumor margins in a rat (2009) Intraoperative fluorescence staining of malignant brain
glioma model. Surg Neurol 40:451 456 tumors using 5 aminofluorescein labeled albumin. Neurosurgery
4. Moore GE (1947) Fluorescein as an agent in the differentiation of 64:53 60, discussion 60 61
normal and malignant tissues. Science 106:130 131 13. Orringer DA, Koo YE, Chen T, Kopelman R, Sagher O,
5. Moore GE, Peyton WT et al (1948) The clinical use of fluorescein Philbert MA (2009) Small solutions for big problems: the applica
in neurosurgery; the localization of brain tumors. J Neurosurg tion of nanoparticles to brain tumor diagnosis and therapy. Clin
5:392 398 Pharmacol Ther 85:531 534
Author Index
A F
Acerbi, F., 251 Fahlbusch, R., 9, 207
Albanese, E., 251 Ferroli, P., 251
Antoniadis, G., 17, 107 Fischer, C.M., 191
Arica, O., 55 Fomekong, E., 139
Atsumi, H., 215 Franzini, A., 251
G
B Ganslandt, O., 207
Behr, M., 241 Gardill, A., 107
Belšán, T., 145, 157 Gasser, T., 35, 49, 61, 73
Beneš, V., 145, 157 Greer, A.D., 151, 231
Bergese, S.D., 43
Bernays, R.L., 191
Bertalanffy, H., 191 H
Bijlenga, P., 111 Hadani, M., 29
Bink, A., 49 Hall, W.A., 119
Black, P.M., 3, 207 Hayashi, M., 67
Bootz, F., 237 Heckelmann, M., 49
Bozinov, O., 191 Hedderich, J., 103
Broggi, G., 251 Heinen, C.P.G., 107
Broggi, M., 251 Hernesniemi, J., 247
Buchfelder, M., 207 Hlavac, G., 17
Bucholz, R.D., 223 Huang, D L., 259
Burkhardt, J K, 191
I
Ikuta, S., 67
C Imoehl, L., 73
Chen, T., 259 Inoue, G., 215
Chernov, M., 67 Iseki, H., 67
Chicoine, M.R., 97
Chiocca, E.A., 43 J
Coenen, V.A., 187 Jolesz, F.A., 3, 21, 207
Cosnard, G., 139
K
D Kapapa, T., 107
Dacey, R.G., 97 Kenmochi, I., 219
Dashti, R., 247 Kiris, T., 55
Dawirs, S., 103 Kockro, R.A., 191
de Rochemont, R du M, 73 Koizumi, J., 215
Dörner, L., 103 König, R.W., 17, 107
Dowling, J.L., 97 Kopelman, R., 259
Kotowski, M., 111
E Kramár, F., 145, 157
Eichhorn, K.W.G., 237 Kretschmer, T., 107
Evans, J.A., 97 Kuhnt, D., 207
265
266 Author Index
L S
Laakso, A., 247 Sagher, O., 259
Lang, M.J., 151, 231 Sandalcioglu, I.E., 61
Laycock, K.A., 223 Santiago, P., 97
Leonard, J.R., 97 Schaller, K., 111
Leuthardt, E.C., 97 Schichor, C., 163
Lim, C.C.H., 97 Schils, F., 175
Limbrick, D.D., 97 Schmidt, T., 17, 107
Lindseth, F., 181 Schnell, O., 163
Schulder, M., 81
M Schwartz, F., 103
Marquardt, G., 73 Seifert, V., 35, 49, 61, 73
Martin, X.P., 139 Selbekk, T., 181
Maruyama, T., 67 Senft, C., 35, 49, 61
Masopust, V., 157 Shariat, K., 169
Matsumae, M., 215 Shinohara, C., 67
McDurmont, L., 223 Singla, A., 97
Medani, K., 3 Smyth, M.D., 97
Mehdorn, H.M., 103 Solheim, O., 181
Meyer, B., 241 Spiro, D., 81
Morhard, D., 163 Steudel, W I, 169
Moriarty, T.M., 89 Stoffel, M., 241
Morikawa, E., 219 Stüer, C., 241
Muragaki, Y., 61, 67 Sure, U., 61
Sutherland, G.R., 151, 231
N Suzuki, T., 67
Nabavi, A., 103 Szelenyi, A., 61
Nabhan, A., 169
Nakajima, Y., 219 T
Netuka, D., 145, 157 Takakura, K., 67
Ng, I., 199 Tanaka, M., 67
Niemelä, M., 247 Titsworth, W.L., 89
Nimsky, C., 61, 207 Tominaga, J., 215
Nishiyama, J., 215 Tonn, J.C., 163
Tringali, G., 251
O Truwit, C.L., 119
Okada, Y., 67 Tsugu, A., 215
Orringer, D.A., 259
Ostrý, S., 145
U
P Uhl, E., 163
Pamir, M.N., 131 Unsgård, G., 181
Pedro, M.T., 107
Pereira, V., 111 V
Peruzzi, P., 43 Vaz, G., 139
Philbert, M.A., 259
Porras, M., 247 W
Puente, E.G., 43 Wirtz, C.R., 17, 107
R Y
Raftopoulos, C., 139 Yoshimitsu, K., 67
Reinke, A., 241 Yoshiyama, M., 215
Reiser, M., 163
Rich, K.M., 97 Z
Ringel, F., 241 Zausinger, S., 163
Rohde, V., 187 Zipfel, G.J., 97
Rüfenacht, D., 111
Subject Index
A Computer navigation kyphoplasty, 177

Advanced Multimodality Image Guided Operating (AMIGO) suite, Coomassie blue (CB), 261
23 24 Cranial window model, 262
AIRIS II MRI scanner, 36 Cystic astrocytoma, 91
American Society of Anesthesiologists (ASA) guidelines, 47
5 Aminolevulinic acid (5 ALA), 67 68 D
Anaplastic ependymoma (WHO III), 90 91 DICOM standard, 224
Aneurysm clipping, 218 Diffusion tensor imaging (DTI), 64, 127, 132, 209
Arterial feeders, 249 Diffusion weighted imaging (DWI), 120, 126, 132
DTI. See Diffusion tensor imaging
B Dual independent twin room 3.0 T. ioMRI system, 143
Bone fiducial markers, 209 DWI. See Diffusion weighted imaging
Brain activation, 120, 126, 127
BrainLAB neuronavigation system, 113 E
Brain mapping, 64 Echo planar imaging (EPI), 62, 127
Brain shift, 37
BrainSuite1 ioMRI Miyabi 1.5 T environment F
ACOM aneurysm, 107 Finite element model (FEM), 239
aneurysm clipping, 110 Flat panel system, intraoperative angiography
feasibility, 108 aneurysm/inadvertent clipping, 114
image guidance, 107, 110 equipment, 113
materials and methods, 107 108 indocyanine green videoangiography, 114
MR TOF angiography, 108 109 infrastructure, 112
patient outcome, 108 intra axial brain tumour surgery, 111
PWI, 109 110 intra operative angiographic control, 111, 113
Brainsuite network system, 200 201, 203 joint neurosurgical neuroradiological intervention, 111, 112
Brain tumor window (BTW) model locoregional and reversible decubitus, 114
applications, 262 263 neuroradiological patients, 113
characterization, 261 262 open neurosurgical patients, 113 115
cranial window, 259, 262 spatial resolution, 7
glioma treatment, 262 vascular structure imaging, 111, 114
materials and methods Fluorescent protoporphyrine IX, 188
caudal most dura, 260 3D Fluoroscopy, 199
coomassie blue (CB), 261 Frameless stereotaxy, 81, 83
craniectomy, 260 Functional endoscopic sinus surgery (FESS), 237 239
9L gliosarcoma cell line, 259
periosteum, 260 G
Tissue Tek, 261 Generalized tonic clonic (GTC) seizure, 44
technique development, 261 Glioblastoma, 55
tumor delineation, 262 Glioma surgical extent of resection, interim analysis
patient demographics, 50, 51
C statistical analysis, 50
Cavernoma, 183, 192, 195 treatment procedure, 50
Central Military Hospital (CMH), 145 Grade II insular astrocytoma, 55
Central nervous system (CNS) tumors, 251 256
Choroidal microcirculation, 251 H
Computed tomography (CT), 215 Heidelberg concept, 18
Computerized tomography, 255 Hemangioblastomas, 255
267
268 Subject Index
HGG. See High grade gliomas hybrid ioMRI suite setup, 200, 202
High field ioMRI iCT scan, 200, 202 203
glioblastoma surgery line of sight neuronavigation system, 204
ALA fluorescence, 104, 105 materials and methods, 199 200
Kaplan Meier survival curves, 105 neuro oncology, 204
KPS, 104 transphenoidal resection, 204
malignant tumor treatment, 105 Interfacing infrared system, 244
microsurgical and neuronavigation (BrainLAB) facilities, 103 Intraoperative computed tomography (iCT)
preop T2fast images, 104 evaluation of imaging, 165
transsphenoidal surgery, 103 ICG, 166
WHO grad IV and III glioma, 103, 104 iCTA and PCT, 164
implementation and preliminary clinical experience intraoperative spinal imaging, 166
aspiration/biopsy, 99 neuronavigation device, 164
brain metastases, 99, 101 radiation exposure, 165
carotid artery, 101 radicality rate, 163
chiari malformations, 99 scanning, 111
epilepsy resections, 99 40 sclice CT scanner, 164
frameless stereotaxy, 97 spine surgery
gliomas, 99, 101 advantage and disadvantage, 173
IMRIS 2 room ioMRI model, 97, 98, 100 complication rate, 171
installation and integration, 98 99 CT suite, 170
maximal safe resection, 101 decompression and stabilization, 169
neurosurgical operating rooms, 97 degenerative and idiopathic listhesis, 170
pituitary adenoma, 99 implants characteristics, 170, 171
retrospective analysis, 98 occipital cervical fixation, 172, 173
time intervals and workflow, 100 patients, 170
wound infections and safety, 99 100 radiation exposure, 173
at 1.5 tesla, 18 19 sacral neurinoma, 172, 173
High grade gliomas (HGG) Somatotom, 170
high field ioMRI, 103 105 spinal instrumentation, 169, 171, 173
iCT, 163 vascular neurosurgery, 165 166
low field ioMRI, 55 59 work flow, 165
neurosurgery, 183, 186 Intraoperative computed tomography angiography (iCTA), 164
Human machine interface (HMI), 231, 232 Intraoperative 3D ultrasound
advantages and disadvantages, 194 195
I image guided neurosurgery, 191 192
ICG VA. See Indocyanine green video angiography image quality, 195
iCT. See Intraoperative computed tomography real time 3D imaging, 193 194
ifMRI. See Intraoperative functional magnetic resonance imaging resection control
ioMRI/ioMRI. See Intraoperative magnetic resonance imaging computerized tomography (CT), 187
Indocyanine green video angiography (ICG VA) craniotomy, 188
cerebrovascular surgery fluorescent protoporphyrine IX, 188
brain arteriovenous malformations, 249 patients and methods, 188, 189
intracranial aneurysms, 248 249 resection cavity, 189 190
intraoperative angiography, 247, 248 tumour remnants, 188 189
microneurosurgical management, 247 Intraoperative electrophysiological monitoring (IOM)
microvascular Doppler, 247, 248 clinical study, 63, 64
NIR video camera, 247 functional and anatomical information, 62
CNS tumors neurophysiological integrity, 61
arteriovenous shunting, 255 phantom study, 63, 64
characteristics, 253, 255 T1 weighted and T2 weighted axial sequences, 62 63
craniotomy, 255 Intraoperative fluorescence angiography (ICG), 166
cystic metastasis, 253, 254 Intraoperative functional magnetic resonance imaging (ifMRI)
patients and methods, 251 252 average registration accuracy, 64
post resection arterial phase, 253 microscope based neuronavigation, 63
pre operative late arterial capillary phase, 253 passive functional MR paradigm, 62
pre operative MRI, 252, 254 somatosensory cortex, 63
revascularization, 254 1.5 T scanner, 62
right frontal metastasis, 252, 254 T2* weighted EPI sequence, 62
vascular neurosurgery, 253 Intraoperative magnetic resonance imaging (ioMRI/ioMRI)
vascular physiopathology, 251 applications, 22
venous drainage, 256 awake craniotomy, supratentorial glioma section
Integrated intra operative room design ASA guidelines, 47
brainsuite network system, 200 201, 203 asleep awake asleep (AAA) technique, 46
casemix, 201, 204 cortical mapping, 45
3D fluoroscopy, 199 dexmedetomidine, 46
Subject Index 269
FLAIR hyperintense abnormality, 45 neuronavigation, 17 19

gross total resection, 47 neurosurgery
GTC seizure, 44 AMIGO suite, 23 24
neuro oncologic surgery, 43 benefits and capabilities, 22
stereotactic navigation, 45 brain shift monitoring methods, 23
tumor debulking, 44 development, 23
brain biopsy, 123 124 imaging tool validation, 23
brain shift, 3, 123 ioMRI applications, 22
coronal T1 weighted contrast enhanced MRI scan, 128 tumor control method, 22
craniotomy, tumor resection, 124 126 patient safety improvement
degree of resection, 3, 6 MRXO, 219
diagnostic 3T scanner, 127 on duty safety nurse, 219, 220
diffusion and perfusion imaging guided tumor resection, 127 psychological factors, 221
divisions of, 17 safety protocol, 221
electric signa SP double coil 0.5 tesla MRI scanner, 120 surgical procedure, 222
Fahlbusch, Rudolph surgical safety check list, 221, 222
advantages and disadvantages, 15 surgical safety manual, 220 221
Black, Peter, 10, 15 World Health Organization, 221, 222
endocrinological methods, 9 pediatric neurosurgery
functional and metabolic imaging, 14 advantages, 93
5 Gauss line, 12, 14, 15 cyst management and CSF diversion, 92 93
INI BrainSuite, 12, 14 historical use, 90
interdisciplinary Neurocenter, 12, 13 low, mid, and high field systems, 89 90
intraoperative CT (Dade Lundsford), 10 in tumor, 90 91
low vs. high field strength, 11 reasons for neurosurgery, 21
medical process optimizing, 15 recurrent gliomas, 5, 6
Odin Pole Star N10, 12 recurrent pituitary adenomas, 6
Peter Heilbrunn’s pilot system, 9 scalpel blade, 128
pilot microscope (Marcovic and Luber), 10 steriotactic biopsy
proton spectroscopy, 14 advantages, 83, 86
Stealth navigation system, 9, 10 computed tomography (CT), 81
Surgiscope (Benabid, Alim Louis and Rose, frameless stereotaxy, 81, 83
Christian Saint), 10 hemiparesis, 83, 85
0.36 TMRI (Koivokangas, John), 12 high and low grade glioma, 83
1.5 T MRI Siemens Sonata, 12, 13 high field ioMRI approach, 86
0.2 T open MRI, 10 Navigus guide, 86
ultrasound technology, 15 patient data, 82 83
volumetric stereotaxy (Patrick Kelly), 9 right thalamic glioma, 83, 84
Zeiss MKM, 10, 11 surgical technique, 82
Zeiss NC4, 11, 12 Sylvian vessels, 83, 85
functional mapping surgical indications, 122 123
DTI, 64 titanium aneurysm clips, 128
functional neuronavigation, 61 3.0T moveable magnet
ifMRI and IOM, 62 64 accurate craniotomy placement, 153, 155
neurophysiological integrity, 61 aneurysm clipping, 153, 155
neurophysiological monitoring, 65 cardiovascular co morbidity, 155
technological pathways, 61 62 clinical material, 152
functional MRI guided tumor resection, 126 127 image acquisition, 155
GE Signa MRT system, 4 5 imaging sequences, 152, 153
glioma surgical extent of resection intracranial neoplasm resection, 153, 154
method, 49 50 magnet and gradients, 151 152
patient demographics, 50, 51 moveable 1.5 tesla (1.5T) magnet, 151
statistical analysis, 50 MR compatible robotics, 156
treatment procedure, 50 operating table, 152
tumor resection, 51, 52 pre operative tractography, 152, 154
grant funding, 6 principles, axiomatic prior, 154
high field ioMRI systems, 121 122 RF coils and shielding, 152
high field experience, 1.5 Tesla, 18 19 sequence acquisition, 152, 153
high grade gliomas, 5 surgical corridor orientation, 155
integration, 17 two room concept, 154
intraoperative complication, 3 Intra operative robotics, neuroArm
low field experience, 0.2 Tesla, 18 clinical application, 233
low field ioMRI systems, 120 121 combinatorial explosion, 234
low grade gliomas, 5, 6 magnet isocenter, 234
magnetic field strength, 119 materials and methods
mid field ioMRI systems, 121 human and machine integration, 232
270 Subject Index
microsurgery, 232, 233 clinical design goals, 29

MR compatible manipulators, 231 physical parameters, 29 30
MRI display, 233 standard operating room, 29
neurosurgical intervention, 235 at 0.2 tesla, 18
non linear processing, 231 Low grade ganglioglioma (WHO I), 90
pre clinical trials, 233, 234 Low grade gliomas (LGG)
robotic surgery, 234 application, 134
shield penetration, 233 multicenter trials, 131
small scale anatomical variability, 234 235 neurosurgery, 183, 185
Intra operative 3.0 tesla magnetic resonance imaging triplanar very high resolution T2W images, 133
average times, 141 143 1.5T technology, 132
blocked transfer table, 140
dual independent twin room 3.0 T. ioMRI system, 143 M
ferromagnetic instrumentation, 140 Magnetic resonance tomography (MRT), 4 5
materials and methods, 139 140 Magnetic resonance/x ray/operation suite (MRXO), 219
MRI related technical problems, 140 141 Magnetoencephalography (MEG), 207
MRI room access, 141, 142 Microneurosurgical clipping, 248
non functional coil problems, 142 Motor evoked potentials (MEPs), 61
pneumencephaly, 142 Multifunctional surgical suite (MFSS). See 3T ioMRI systems
IOM. See Intraoperative electrophysiological monitoring Multimodality imaging suite
ideally positioned radiological equipment, 215 217
K interventional radiology procedures, 216, 217
Karnofsky performance score (KPS), 104 partial splenic embolization, 217
Kolmogorov Smirnov’s test, 50 preoperative planning, 218
sharing imaging equipment, 218
L supratentorial/infratentorial glioma, 216 217
Laser Range Scanner (LRS), 23 twin operating theater, 217
LGG. See Low grade gliomas Multimodal navigation
Low field ioMRI automatic registration, 208
glioma surgery brain shift, 209
development, 35 36 double doughnut GE scanner, 208
image quality, 36 37 frameless stereotaxy, 207
indications, 37 functional navigation, 207, 208
influence of, 37 39 glioma resection, 211
intracranial surgery, 39 MEG, 207
vs. intraoperative ultrasound or intraoperative computed microscope based image injection, 209
tomography, 35 multimodal data integration, 208 209
low vs. high field systems, 39 non linear registration techniques, 211
microneurosurgical techniques and refined imaging pilocytic asrocytoma, 210 211
modalities, 35 Multiplanar reconstructions (MPR), 164
high grade gliomas Multislice (multi detector row) systems (MSCT), 166
clinical features, 58
histopathological examination, 57 N
microsurgical resection, 55 N acetyl aspartate (NAA), 124
patients and method, 55 57 Near infrared (NIR) video camera, 247
Polestar N 20 system, 58, 59 Neo futuristic diagnostic imaging operating suite.
seeding metastasis, 58 See Multimodality imaging suite
Signa Sp/I, 59 Neuronavigation system, 17 19, 145
indications for, 32
information guided surgical management, gliomas O
advantages, 70 O arm guided balloon kyphoplasty
brain mapping, 69, 70 C arm radiological exposure, 175
glioma resection, 70 71 computer navigation kyphoplasty, 177
haemorrhage, 68 fluoroscopy time, 176, 177
ioMRI investigations, 68 69 immediate 3D acquisition, 177
insular glioblastoma multiforme, 69, 71 mean irradiation dose, 177
intelligent operating theatre, 68 osteopenia evaluation, 176
intraaxial brain tumor, 67 procedure and population, 176
intraoperative integration, 70 X ray exposure, 175, 177
postoperative neurological morbidity, 68, 69 Ojemann stimulator, 45
retrospective analysis, 71 72 Open ioMRI scanner, AIRIS IITM, 68
Sylvian fissure baseline iMR images, 69 Operating room integration and telehealth
WHO grade II, III, and IV gliomas, 71, 72 computer based imaging, 223
PoleStar scanner DICOM standard, 224
advantages, 32 neurosurgical procedure, 223 224
clinical application, 31 requirements, 224
Subject Index 271
surgical communication network integration cranio cervical, atlanto axial and subaxial pathologies, 242
device control, 225 neurological damage, 242
remote access, 226 operation steps, 242
SurgON, 224, 226 paradigm shift, 241
unified multi source display system, 225 preliminary results, 242
relative position, 242, 244
P SpineAssist1, 241, 242
Parkinson’s disease, 223 total average position deviation, 242, 243
Perfusion computed tomography (PCT), 164 in vivo spinal procedures, 244
Perfusion weighted imaging (PWI), 109 110
Peritumoral brain edema, 123 S
Personal data assistant (PDA), 225 Sensory evoked potentials (SEPs), 61
Photon radiosurgery system (PRS), 68 Siemens 1.5 T Brain Suite design, 132
Pilocytic asrocytoma, 210 211 Signa Sp/I, 59
Pituitary adenomas Single voxel spectroscopy (SVS), 124
resection control, 74 78 Skull base tumors, 183
3T ioMRI, Istanbul experience, 131, 133, 135 Sonography, 195
PoleStar ioMRI system, 36 SonoWand1, 181
PoleStar N20 SpineAssist1, 241, 242
cranial surgery, 58 Splenomegaly, 217
high grade gliomas, 59 Statistical parametric mapping (SPM), 62
interim analysis, 50 Stereo camera imaging method, 23
resection control, pituitary adenomas Stereotactic brain biopsy, 81
coronal T1 weighted gandolinium DTPA, 75 Supplementary motor area (SMA) syndrome, 45
disadvantages, 78 Surgical operative network (SurgON), 224, 226
ioMRI developmental trend, 77 78 Surgical Planning Lab (SPL), 4
navigation guided resection, 78
ophthalmologic status, 77 T
patient positioning, 74 3T ioMRI systems
patients and methods, 74 76 Acibadem University ioMRI suite
pituitary macroadenomas, 77 image quality and capabilities, 133
standard microsurgical procedure, 74 low grade glioma surgery, 134 135
tumor resection, 75, 76 shared resource Siemens Trio 3T system, 132, 133
stereotactic biopsy, 82, 83, 86 transsphenoidal surgery, 134 136
PoleStar scanner twin room/shared resource design, 133 134
advantages, 32 MFSS
clinical application, 31 acute ischaemia, intraoperative assessment, 147
clinical design goals, 29 biopsy, 147
physical parameters, 29 30 Central Military Hospital, 145
standard operating room, 29 extradural spinal tumour, 147
Polyetheretherketone, 232 imaging quality, 146
Portal hypertension, 217 immediate postoperative imaging, 146 147
Posterior fossa pilocytic astrocytoma, 91 intended partial resection, 147
Preoperative inspection, 219 intraoperative navigation, 147
Programmable universal machine for assembly (PUMA), 231 materials and methods, 145 146
Prospective stereotaxy, 123 multicystic and cystic lesions, 147
pituitary adenomas, 146
R resection extension, epilepsy surgery, 147
RASS. See Robot assisted endoscopic sinus surgery skull base and spinal cord tumour, 147
Registration matrix, 208 spine surgery, 147 148
Remote station, 226 neurooncology, 131, 132
Robot assisted endoscopic sinus surgery (RASS), 237, 240 pituitary adenoma
Robot assisted endoscopy bilateral endonasal approach, 157
clinical requirements, 237 hormone normalization, 159
DaVinci System, 239 low and high field MRI scanners, 157
force/torque senor, 238 239 microadenomas and macroadenomas, 158
geometrical centre, 238 non pituitary adenoma lesions, 158
methods/material, 238 pituitary adenoma resection, 157, 158
otorhinolaryngology, 237 radical resection, 158, 159
sinus surgery modelling, 238 suprasellar growth, 158
technical development, 239 proton MRS, 132
translations and rotations, 240 Signa SP system, 132
Robotic assisted stereotaxy, 233 234 0.5 T General Electric magnet, 131
Robotic technology, spine surgery 1.5T technology, 132
autonomous robots, 244 Total intravenous anaesthesia (TIVA), 62
computer assisted surgical equipment, 244 245 Tractography images, 200
272 Subject Index
Turbo spectroscopic imaging (TSI), 124 image guided resection, 182 183
T2* weighted EPI sequence, 62 image quality, 182, 185
Twin operating theater, 217 resection control, 183
SonoWand1, 181
U tumour delineation, 182
Ultra high field MRI Ultrasound angiography
DTI, 132 real time 3D ultrasound imaging, 194
low grade glioma patients, 134, 135 3D ultrasound assisted image guided neurosurgery, 192
shared resource Siemens Trio 3 T system, 133 Uterine cervical cancer, 217
3T scanners, 132
Ultra low field ioMRI. See Low field ioMRI; Polestar N20 V
3D Ultrasound Ventricle catheter, 184
acquisition and display Vertebral compression fractures. See O arm guided balloon
image guided neurosurgery, 192 kyphoplasty
real time 3D ultrasound imaging, 193 194
neurosurgery W
applications, 183 186 Wide area network (WAN), 226
brain shift, 181 Wilcoxon Mann Whitney U test, 50
2D US, transsphenoidal approach, 183, 184 World Health Organization (WHO), 221, 222

Intraoperative Imaging PP 169-174

Uploaded by

Copyright:

Available Formats

Intraoperative Imaging PP 169-174

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Intraoperative Imaging PP 169-174

Uploaded by

Copyright:

Available Formats

FHB

(Radiology golden books)

Editor: H.-J. Steiger

This work is subject to copyright.

Typesetting: SPI, Pondichery, India

Printed on acid free and chlorine free bleached paper

Library of Congress Control Number: 2010935184

With 117 (partly coloured) Figures

ISSN 0065 1419

In the pursuit of the goal of continuous improvement in surgical results, intraoperative

Istanbul, Turkey M.N. Pamir, T. Kırış

History- Development- Prospects of Intraoperative Imaging

From Vision to Reality: The Origins of Intraoperative MR Imaging . . . . . . . . . . . . . . . . . . 3

Development of Intraoperative MRI: A Personal Journey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Lows and Highs: 15 Years of Development in Intraoperative

Intraoperative Imaging in Neurosurgery: Where Will the Future Take Us? . . . . . . . . . . . 21

Intraoperative MRI- Ultra Low Field Systems

Development and Design of Low Field Compact Intraoperative MRI

Low Field Intraoperative MRI in Glioma Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

Intraoperative MRI (ioMRI) in the Setting of Awake Craniotomies

Glioma Extent of Resection and Ultra-Low-Field ioMRI: Interim Analysis

Impact of a Low-Field Intraoperative MRI on the Surgical Results

Intraoperative MRI and Functional Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Information-Guided Surgical Management of Gliomas Using

Implementation of the Ultra Low Field Intraoperative MRI PoleStar

Intraoperative MRI for Stereotactic Biopsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

The Evolution of ioMRI Utilization for Pediatric Neurosurgery:

Intraoperative MRI - High Field Systems

Implementation and Preliminary Clinical Experience with the

High-Field ioMRI in Glioblastoma Surgery: Improvement

Image Guided Aneurysm Surgery in a Brainsuite1 ioMRI Miyabi

From Intraoperative Angiography to Advanced Intraoperative Imaging:

Intraoperative MRI - Ultra High Field Systems

Intraoperative Magnetic Resonance Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

3 T ioMRI: The Istanbul Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

Intra-operative 3.0 T Magnetic Resonance Imaging Using a

Multifunctional Surgical Suite (MFSS) with 3.0 T ioMRI: 17 Months

Intra-operative MRI at 3.0 Tesla: A Moveable Magnet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

Intraoperative CT and Radiography

Intraoperative Computed Tomography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

Intraoperative CT in Spine Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

O-Arm Guided Balloon Kyphoplasty: Preliminary Experience

Intra-operative Imaging with 3D Ultrasound in Neurosurgery . . . . . . . . . . . . . . . . . . . . . . . . 181

Intraoperative 3-Dimensional Ultrasound for Resection Control During Brain

Advantages and Limitations of Intraoperative 3D Ultrasound

Integrated Intra-operative Room Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199

Multimodal Navigation Integrated with Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

Multimodality Imaging Suite: Neo-Futuristic Diagnostic Imaging Operating

Improving Patient Safety in the Intra-operative MRI Suite Using

Operating Room Integration and Telehealth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223

Other Intraoperative Imaging Technologies and Operative Robotics

Intra-operative Robotics: NeuroArm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231

Clinical Requirements and Possible Applications of Robot Assisted

Robotic Technology in Spine Surgery: Current Applications

Microscope Integrated Indocyanine Green Video-Angiography

Application of Intraoperative Indocyanine Green Angiography

A Technical Description of the Brain Tumor Window Model:

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267