Modeling brain connectivity from multi-modal imaging data
- Preliminary topic: Modeling brain connectivity from multi-modal imaging data
- Start: As soon as possible
- Place: Dept. of Nuclear Medicine, Klinikum rechts der Isar
- Supervisor: Igor Yakushev, MD (Nuclear Medicine) and Dr. Kuangyu Shi, PhD? (CAMP)
Topic
There are a few neuroimaging techniques that provide useful information about brain structure and function. For instance, positron emission tomography (PET) measures neuronal metabolism, magnetic resonance imaging (MRI) measures the thickness of brain cortex, while diffusion tensor imaging (DTI) allows reconstruction of neural tracts connecting different brain regions.
A modern hybrid PET/MR system at the Dept. of Nuclear Medicine enables acquisition of PET, MRI and DTI data in the same subject at one haul. Our recent, successful work in co-operation with CAMP revealed that connectivity analyses of PET data are superior to traditional analytical techniques in respect to differential diagnosis of dementia disorders. The aim of this work is to develop algorithms that allow integration of connectivity information from a few imaging modalities. It should be further tested if this approach improves accuracy to classify different forms of dementia.
Conditions
We offer an intensive supervision, but initiative is very welcome. Working environment will be provided at the Dept. of Nuclear Medicine at the Klinikum rechts der Isar (Ismaninger Str. 22, 81675 München).
We offer a
HiWi? support of 450 EUR (net) per month after a probe period of one month.
Requirements
The student should be interested in medical imaging and image processing. Knowledge of Matlab is required. A strong mathematical background in linear algebra or differential equations is advantageous.
Contact
This project is jointly proposed between the Chair for Computer Aided Medical Procedures (Prof. Nassir Navab, I16, Department of Informatics, TUM) and the Dept. of Nuclear Medicine (Prof. Markus Schwaiger, Klinikum rechts der Isar, TUM).
Email addresses of former students (for references) can be obtained upon request without obligation.