CN106558045B - A kind of segmentation of lung parenchyma method, apparatus, magic magiscan - Google Patents
A kind of segmentation of lung parenchyma method, apparatus, magic magiscan Download PDFInfo
- Publication number
- CN106558045B CN106558045B CN201610912654.0A CN201610912654A CN106558045B CN 106558045 B CN106558045 B CN 106558045B CN 201610912654 A CN201610912654 A CN 201610912654A CN 106558045 B CN106558045 B CN 106558045B
- Authority
- CN
- China
- Prior art keywords
- image
- medical image
- connected domain
- segmentation
- cross
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10088—Magnetic resonance imaging [MRI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10104—Positron emission tomography [PET]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30061—Lung
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The invention discloses a kind of segmentation of lung parenchyma methods, comprising: obtains the first medical image of subject target area, the first medical image includes several slice images, and each slice image includes multiple pixels；The sagittal plane of first medical image is linearly enhanced, and the first medical image is added with the gray value by the first image respective pixel linearly enhanced, generates the second medical image；It determines the sagittal plane skin line of the second medical image, and determines the cross section skin line of the second medical image on the second medical image that sagittal plane skin line determines；The connected domain that sagittal plane skin line and cross section skin line surround is obtained, and obtains the segmentation result of lung tissue in connected domain.Segmentation of lung parenchyma method of the present invention can accurately obtain the borderline region of lung tissue.Meanwhile the present invention also proposes a kind of segmentation of lung parenchyma device and the magic magiscan using the segmentation of lung parenchyma device.
Description
Technical field
The present invention relates to the dividing methods of lung tissue, dress in technical field of medical image processing more particularly to medical image
It sets and magic magiscan.
Background technique
Prostate specific antigen instrument (Emission Computed Tomography, ECT) mainly includes monochromatic light
Sub- emission computed tomography (SPECT) and PET-Positron emission computed tomography scanning (PET), wherein PET is as current
High-level nuclear medicine technology, it has also become the indispensable important method that tumour, the heart, cerebral disease diagnose.Magnetic resonance imaging
(Magnetic Resonance Imaging, MRI) can provide the anatomic form and physiological function information at imaging position, react
It is preferably right that there is unrivaled superiority, especially soft tissue MR image to present for anatomic form and physiological function information approach
Than effect, and it is radiationless.The combination that PET and MR is checked has many advantages, such as that sensitivity is good, accuracy is high.PET/MR it is multi-modal at
As in system, needing to carry out PET image correction for attenuation, but due to MR image pixel value only in tissue proton density and
The relaxation degree of tissue is related, and mass attentuation coefficient relevant to electron density is unrelated, such as bone and air are respectively provided with
Highest and lowest positive electron attenuation coefficient is but all low signal on MR image, so the correction for attenuation ratio based on MR image
It is more difficult.
Region segmentation method is broadly divided into currently based on the attenuation correction method of MR: attenuation characteristic different tissue and device
Official is divided into different regions, such as air, lung, fat, muscle and bone, then obtains the different zones divided again
The attenuation coefficient of respective organization under 511keV carries out correction for attenuation.In the prior art, the precision of the method based on region segmentation has
Wait further increase, and it cannot achieve the segmentation of different tissues and organ in more bed magnetic resonance image.In consideration of it, it is necessary to right
Existing medical image cutting method improves.
Summary of the invention
The technical problem to be solved by the present invention is to propose that a kind of precision is high and can realize the medical image of more bed scannings
Dividing method.
The technical scheme of the invention to solve the technical problem is: a kind of segmentation of lung parenchyma method, including it is as follows
Step:
The first medical image of subject target area is obtained, first medical image includes several slice images,
Each slice image includes multiple pixels；
The sagittal plane of first medical image is linearly enhanced, and to first medical image and by linear
The gray value of first image respective pixel of enhancing is added, and generates the second medical image；
Determine the sagittal plane skin line of second medical image, and the second medicine determined in the sagittal plane skin line
The cross section skin line of second medical image is determined on image；
The connected domain that the sagittal plane skin line and cross section skin line surround is obtained, and obtains lung in the connected domain
The segmentation result of tissue.
Optionally, first medical image is CT image or MR image.
Optionally, the sagittal plane skin line of second medical image is obtained by following process:
Binary conversion treatment is carried out using given threshold to each pixel of the second medical image sagittal plane；
Count the quantity of the included connected domain of each lamella of the second medical image sagittal plane after binary conversion treatment；
Judge that the quantity of the included connected domain of each lamella whether less than the first given threshold, if the conditions are met, then retains
The connected domain；Otherwise, the only maximum one or more connected domains of Retention area in included connected domain；
The sagittal plane skin line of second medical image is determined according to the connected domain of the reservation.
Optionally, further include obtain slice image boundary on pixel value be 1 point, and according to the pixel value be 1 point
The connected domain of reservation is closed.
Optionally, the cross section skin line of second medical image is determined by following process:
Obtain the second medical image that the sagittal plane skin line determines；
Two-dimentional corrosion treatment is carried out to each lamella in the second medical image cross section that the sagittal plane skin line determines,
Retention area is greater than the connected domain of the first setting area threshold；
The connected domain for being greater than the first setting area threshold to the area of the reservation carries out two-dimensional swelling processing, and according to two
Connected domain after dimension expansion process determines the cross section skin line of second medical image.
Optionally, the segmentation result of the lung tissue is obtained by following process:
The boundary pixel point of high gray value connected domain is obtained in the cross section of the connected domain；
Cross section segmented image and coronal-plane segmented image are obtained in the connected domain that the boundary pixel point surrounds, it is described
Cross section segmented image or the coronal-plane segmented image are low ash angle value region；
Center line is obtained according to the coronal-plane segmented image；
Lung group is determined in the position of the cross section segmented image and the coronal-plane segmented image according to the center line
Knit corresponding connected domain.
Optionally, the center line is obtained by following process:
The connected domain area for calculating each lamella of coronal-plane segmented image, determines the maximum lamella of connected domain area；
The initial row and termination row of connected domain are determined in the maximum lamella of connected domain area；
Center line is determined according to the initial row of the connected domain and termination row.
Optionally, true in the position of the cross section segmented image and the coronal-plane segmented image according to the center line
Determine the specific steps of the corresponding connected domain of lung tissue are as follows:
Retain center line connected domain of the row in the coronal-plane segmented image；
Center line connected domain of the row is filled in the cross section segmented image, obtains lung tissue pair
The connected domain answered.
According to another aspect of the present invention, it is also proposed that a kind of segmentation of lung parenchyma device, including image segmentation module, the image
Dividing module includes:
First medical image acquisition unit, for receiving the first medical image of subject target area, first doctor
Learning image includes several slice images, and each slice image includes multiple pixels；
Second medical image generation unit is linearly enhanced for the sagittal plane to first medical image, and right
First medical image is added with the gray value by the first medical image respective pixel linearly enhanced, generates the second medicine
Image；
Skin line determination unit, for determining the sagittal plane skin line of second medical image, and in the sagittal plane
The cross section skin line of second medical image is determined on the second medical image that skin line determines；
Cutting unit, the connected domain surrounded for obtaining the sagittal plane skin line and cross section skin line, and described
The segmentation result of lung tissue is obtained in connected domain.
According to another aspect of the invention, it is also proposed that a kind of magic magiscan, the magic magiscan
Including segmentation of lung parenchyma device, further includes:
MR scan module for scanning subject target area, and obtains the corresponding first medicine figure in the target area
Picture；
PET scan module for scanning the target area, and acquires the corresponding PET data in target Europe region；
Module is rebuild, for obtaining the segmentation of lung parenchyma of the segmentation of lung parenchyma device acquisition as a result, and being the segmentation
The each pixel of lung tissue distribute respective attenuation coefficient, generate decay pattern, and the PET according to the decay pattern iterative approximation
Data generate PET image.
Compared with the prior art, the advantages of the present invention are as follows: it is cut in conjunction with magnetic resonance image in differences such as sagittal plane, cross sections
The gray scale of face image, location information obtain the connected domain that sagittal plane skin line and cross section skin line surround, and realize medical image
Effective differentiation of the image tissues such as the similar background area of gray value and lung, bone, avoids background area to lung segmentation knot
The influence of fruit；Segmentation obtains high gray value area in the connected domain cross section that sagittal plane skin line and cross section skin line surround
Domain obtains the connected domain that lung tissue may include, and is sieved according to the structure feature selected center line of lung tissue to connected domain
Choosing effectively excludes influence of the bone to lung tissue；Without complicated priori matching template, it is applicable to more bed scannings.
Detailed description of the invention
Fig. 1 is the magic magiscan structural block diagram of one embodiment of the invention；
Fig. 2 is the MR scan module structural block diagram of one embodiment of the invention；
Fig. 3 is the PET scan modular structure block diagram of one embodiment of the invention；
Fig. 4 is the segmentation of lung parenchyma apparatus structure block diagram of one embodiment of the invention；
Fig. 5 is the segmentation of lung parenchyma method flow diagram of one embodiment of the invention；
Fig. 6 a is the cross-sectional view strength for the first medical image that one embodiment of the invention obtains；
Fig. 6 b is the coronal-plane view for the first medical image that one embodiment of the invention obtains；
Fig. 6 c is the sagittal plane view for the first medical image that one embodiment of the invention obtains；
Fig. 7 is that the sagittal plane skin line of the second medical image of one embodiment of the invention obtains flow chart；
Fig. 8 is that one embodiment of the invention obtains lung group in the connected domain that sagittal plane skin line and cross section skin line surround
The segmentation result flow chart knitted；
Fig. 9 a is the cross section skin line result schematic diagram that one embodiment of the invention obtains；
Fig. 9 b is the segmentation of lung parenchyma result schematic diagram that one embodiment of the invention obtains.
Specific embodiment
In order to illustrate more clearly of the technical solution of embodiments herein, will make below to required in embodiment description
Attached drawing is briefly described.It should be evident that the accompanying drawings in the following description is only some examples or implementation of the application
Example, for those of ordinary skill in the art, without creative efforts, can also be according to these attached drawings
The application is applied to other similar scene.Unless being explained obviously or separately, identical label generation in figure from language environment
The identical structure or operation of table.
As shown in the application and claims, unless context clearly prompts exceptional situation, " one ", "one", " one
The words such as kind " and/or "the" not refer in particular to odd number, may also comprise plural number.It is, in general, that term " includes " only prompts to wrap with "comprising"
Include clearly identify the step of and element, and these steps and element do not constitute one it is exclusive enumerate, method or apparatus
The step of may also including other or element.
Although the application is made that various draw to the certain module in data processing system according to an embodiment of the present application
With however, any amount of disparate modules can be used and operate in the client connecting by network with the system
And/or on server.The module is merely illustrative, and different moulds can be used in the different aspect of the system and method
Block.
Flow chart used herein is used to illustrate performed by data processing system according to an embodiment of the present application
Operating procedure.It is not necessarily accurately carried out in sequence it should be understood that being shown in the operating procedure of above or below.Phase
Instead, various steps can be handled according to inverted order or simultaneously.It is also possible to during other operating procedures are added to these,
Or a certain step or number step operation are removed from these processes.
In medical image or data handling procedure, " image segmentation ", " image zooming-out ", " image classification " can mutually turn
Change, the image for meeting certain condition is chosen in expression out of extensive area.In some embodiments, magic magiscan can
By include it is one or more in the form of.The form includes but is not limited to digital subtraction angiography (DSA), magnetic resonance imaging
(MRI), magnetic resonance angiography (MRA), computed tomography (CT), computed tomography angiography (CTA), ultrasound
Wave scans (US), positron emission tomography (PET), single photon emission computerized tomography,SPECT (SPECT), SPECT-
MR, CT-PET, CE-SPECT, DSA-MR, PET-MR, PET-US, SPECT-US, TMS-MR, US-CT, US-MR, X-ray-CT,
X-ray-PET, X-ray-US, video-CT, video-US and/or similar one or more combinations.In some embodiments,
The target area of image scanning can be where one or more combination such as organ, body, object, damage location, tumour
Region.In some embodiments, the target area of image scanning can be thoracic cavity, abdomen, organ, four limbs, bone, blood vessel etc. one
Region where kind or a variety of combinations.In some embodiments, the target area of scanning can be one or more positions
Region where organizing.In some embodiments, image can be two dimensional image and/or 3-D image.In two dimensional image, most
Subtle resolvable elements can be pixel (pixel).In 3-D image, most subtle resolvable elements can be voxel
(voxel).In 3-D image, image can be by a series of two dimension slicing or two-dimensional slice image construction.
It should be noted that the description below for magic magiscan only for convenience of description can not be this Shen
It please be limited within the scope of illustrated embodiment.It is appreciated that for those skilled in the art, in the original for understanding the system
After reason, without departing substantially from this principle, any combination is carried out to modules, or constitute subsystem and other modules
Connection, to the various modifications and variations of the implementation above method and systematic difference field in form and details.
According to some embodiments of the present invention, a kind of segmentation of lung parenchyma device, including image segmentation module, the image are proposed
Dividing module includes: the first medical image acquisition unit, and for receiving the first medical image of subject target area, this first
Medical image includes several slice images, and each slice image includes multiple pixels；Second medical image generation unit, is used for
The sagittal plane of first medical image is linearly enhanced, and to first medical image and the first medicine by linearly enhancing
The gray value of image respective pixel is added, and generates the second medical image；Skin line determination unit, for determining the second medical image
Sagittal plane skin line, and the cross-section of second medical image is determined on the second medical image that sagittal plane skin line determines
Surface skin line；Cutting unit, the connected domain surrounded for obtaining the sagittal plane skin line and cross section skin line, and described
The segmentation result of lung tissue is obtained in connected domain.In one embodiment, the first medical image acquisition unit can be for storage
The memory of function, the memory can be floppy disk, CD, CD-ROM (compact-disc-read-only memory), magneto-optic disk, ROM (only
Read memory), RAM (random access memory), EPROM (Erasable Programmable Read Only Memory EPROM), (electric erasable can by EEPROM
Program read-only memory), magnetic or optical card, flash memory or suitable for store machine-executable instruction other kinds of medium/machine
Readable medium.
In some embodiments, which can be with central processing unit (Central
Processing Unit, CPU), specialized application integrated circuit (Application Specific Integrated
Circuit, ASIC), dedicated instruction processor (Application Specific Instruction Set Processor,
ASIP), physical processor (Physics Processing Unit, PPU), digital signal processor (Digital
Processing Processor, DSP), field programmable gate array (Field-Programmable Gate Array,
FPGA), programmable logic device (Programmable Logic Device, PLD), processor, microprocessor, controller, micro-
The combined processor of one or more of controller etc., for executing aforesaid operations.
In further embodiments, segmentation of lung parenchyma device may include the first medical image generation module, first medicine
Image generating module can be the MR scanner connecting with image segmentation module, and the first medical image generation module can be to being examined
Person target area is scanned, and generates the anatomic image of subject target area.Illustratively, which generates single
It is disconnected that member can be magnetic resonance scanner (equipment), magnetic resonance angiography scanner, computed tomography scanner, computer
Layer vessel scanning angiography scans instrument etc. can produce the imaging device of subject target area anatomical information.
According to some embodiments of the present invention, a kind of Medical Image Processing system using aforementioned segmentation of lung parenchyma device is proposed
System, to handle multi-modality medical image.In one embodiment, which may include segmentation of lung parenchyma dress
It sets, which includes image segmentation module, and image segmentation module can realize the automatic segmentation of lung tissue.?
In another embodiment, which may include image segmentation module and other scan modules.
In some embodiment of the invention, if Fig. 1 is the magic magiscan structural frames of some embodiments of the invention
Figure, which includes MR scan module 100, PET scan module 200, the lung containing image segmentation module 300
Tissue segmentation device 30 rebuilds module 400, control module 500, display module 600.Wherein, the storage in MR scan module 100
Unit 103 is connect with the first medical image acquisition unit 301 of image segmentation module 300；The segmentation list of image segmentation module 300
First 304, the memory 204 of PET scan module 200 is connect with reconstruction module 400 respectively；It is single that control module 500 can connect segmentation
Member 304 rebuilds module 400 or display module 600.
Fig. 2 is 100 structural block diagram of MR scan module of one embodiment of the invention, and illustratively, which can
To be MR scanner, including MR signal generates and acquisition unit 101, MR signal processing unit 102, storage unit 103 and MR control
Unit 104 processed, and can connect between each other, which can be wireless network connection or finite element network connection.
MR signal generates and acquisition unit 101 may include magnet and radio-frequency coil, and magnet includes the master for generating main field
Magnet and the gradient component for generating gradient.Main magnet can be permanent magnet or superconducting magnet；Gradient component mainly includes gradient electricity
Stream amplifier (AMP), gradient coil；Gradient component also may include three autonomous channels Gx, Gy, Gz, and each gradient amplifier swashs
A corresponding gradient coil, generates the gradient fields for generating additional space encoded signal, to magnetic in heat gradient coil group
Resonance signal carries out space orientation；Radio-frequency coil can be divided into radio-frequency sending coil and RF receiving coil, and radio-frequency sending coil is used
In to subject or human-body emitting RF pulse signal, RF receiving coil is used to receive the magnetic resonance signal acquired from human body.
Optionally, the type of radio-frequency coil can be bird basket coil, solenoid-shaped coil, saddle-type coil, Helmholtz coil, battle array
Alignment circle, circuit coil etc..
It is single with acquisition unit 101, MR signal processing unit 102 and storage that MR control unit 104 can control MR signal to generate
Member 103.Specifically, MR control unit 104 is connected with comprising pulse-series generator, gradient waveform generator, transmitter and connects
Receipts machine etc. is receiving user after the instruction that console issues, and control MR signal generates and acquisition unit 101 executes respective scanned
Sequence；MR signal processing unit 102 can receive the magnetic resonance signal that MR signal generates and acquisition unit 101 acquires, and MR is controlled
Unit 104 controls MR signal processing unit 102 and carries out the operation such as Fourier transformation to magnetic resonance signal, generates the magnetic at imaging position
Resonance image；Storage unit 103 can store the magnetic resonance image under the control of MR control unit 104.
Storage unit 103 can make the memory with store function, and the memory includes, but are not limited to floppy disk, light
Disk, CD-ROM (compact-disc-read-only memory), magneto-optic disk, ROM (read-only memory), RAM (random access memory), EPROM
(Erasable Programmable Read Only Memory EPROM), EEPROM (electrically erasable programmable read-only memory), magnetic or optical card, flash memory or
Suitable for storing other kinds of medium/machine readable media of machine-executable instruction.
Illustratively, the first medical image can be MR image, and MR scan module 100 generates the in one embodiment of the invention
The detailed process of one medical image includes: that main magnet generates B0 main field, and the intracorporal atomic nucleus of subject is under main field effect
Precession frequency is generated, the precession frequency and main field strength are proportional；What the storage of MR control unit 104 and transmission needed to be implemented sweeps
Retouch the instruction of sequence (scan sequence), pulse-series generator instructed according to scanning sequence to gradient waveform generator and
Transmitter is controlled, and gradient waveform generator output has the gradient pulse signal of scheduled timing and waveform, which passes through
Gx, Gy and Gz gradient current amplifier, then pass through three autonomous channels Gx, Gy, Gz in gradient component, each gradient amplifier
A corresponding gradient coil in gradient coil set is excited, the gradient fields for generating additional space encoded signal are generated, with right
Magnetic resonance signal carries out space orientation；Pulse-series generator also executes scanning sequence, and output includes the RF pulse-to-pulse of radio-frequency transmissions
The timing of the data such as timing, intensity, the shape of punching and radio frequency reception and the length of data acquisition window are sent out simultaneously to transmitter
It penetrates machine and respective radio-frequency pulse is sent to the generation of MR signal and the generation of acquisition unit 101 B1 comprising radio-frequency sending coil,
The signal that the atomic nucleus being excited in patient body under B1 field action issues is generated and is adopted by the MR signal comprising RF receiving coil
Collection unit 101 perceives；Then, MR signal processing unit 102 is transferred to by transmission/reception switch, by amplification, demodulation,
The digitized processings such as filtering, AD conversion can form raw k-space data, which is fourier transformed and can rebuild
For the first medical image (MR image), and it is stored in storage unit 103.
Such as 200 structural block diagram of PET scan module that Fig. 3 is some embodiments of the invention.The PET scan module, which can be, adopts
Collect the PET scanner of subject target area PET data, it may include detector cells 201, PET signal processing unit 202, symbol
Total counting unit 203, storage unit 204 and PET control unit 205.PET control unit 205 can control other multiple unit works
Make execution probe command, detector cells 201 include the multiple detector rings being arranged on the rack, which has arrangement
Multiple detectors on central axis circumference, subject can be at scan vision (the Field Of surrounded by multiple detectors
View, FOV) in imaging.
Storage unit 204 can be the memory with store function, the memory include, but are not limited to floppy disk, CD,
CD-ROM (compact-disc-read-only memory), magneto-optic disk, ROM (read-only memory), RAM (random access memory), EPROM (can
Erasable programmable read-only memory (EPROM)), EEPROM (electrically erasable programmable read-only memory), magnetic or optical card, flash memory or be suitable for
Store other kinds of medium/machine readable media of machine-executable instruction.
Illustratively, above-mentioned pet scanner obtains the process of PET data are as follows: before PET scan, infuses into subject's body
Enter the medicament (tracer) of radioactive isotope mark；What detector detection was released inside subject penetrates at pair annihilation gamma
Line generates pulse type electric signal corresponding with the light quantity at pair annihilation gamma ray detected；The pulse type electric signal is supplied
To PET signal processing unit 202, which generates single event data (Single according to electric signal
Event Data), PET signal processing unit 202 is more than threshold value this case by the intensity of detection electric signal in practice, thus
Electro-detection annihilation gamma ray；Single event data are supplied to coincidence counting unit 203, the coincidence counting unit 203 to it is multiple
The related single event data of single event are implemented while counting processing, and specifically, coincidence counting unit 203 is supplied from repetition
Determination is repeated in single event data is contained in event data related with two single events in preset time range, when
Between range be set to such as 6ns~18ns or so, the pairs of single event be presumed to origin in from same at pair annihilation point
The pairs of annihilation gamma ray generated, wherein pairs of single event, which is briefly referred to as, meets event, connection detects that this falls into oblivion in pairs
Do not have the line of gamma-ray pairs of detector to be referred to as line of response (Line Of Response, LOR), it is corresponding with line of response is met
The data that meet be PET data；Storage unit 204 can store PET data under the control of PET control unit 205.It needs to illustrate
, since PET scan equipment and measurement process are there are error, which also needs to carry out detector sensitivity correction, same to position
Plain time decay correction, coincidence correction, coincidence correction, scatter correction, correction for attenuation or geometric correction and other corrections.
The medicament of the above-mentioned radioactive isotope mark being related to can be reversible tracer or irreversible tracer, PET are swept
The process retouched can be using single tracer scanning or more tracer dynamic scans.PET scan is double tracers in one embodiment
Agent dynamic scan, the first tracer I comprising initial time injection1With T0Second tracer I of moment injection2, and first shows
Track agent I1For reversible tracer, the second tracer is irreversible tracer, dynamic scan specifically: PET scan since t=0,
And the first tracer I is injected into subject's body in t=01；In t=T0When inject the second tracer I2, by time T1Scanning
Terminate, the entire scanning process duration is T0+T1.During the scanning process, it is carried out using head of the detector to injection tracer
Real-time detection obtains the radiated signal that subject head issues, and meets detection by this and acquisition system is handled, formed original
Meet data, frequency acquisition is that the acquisition of per unit moment is primary, obtains T0+T1Group coincidence counting, in which: in T0In period
The T collected0The corresponding first tracer I of group coincidence counting1, in T1The T collected in period1Group coincidence counting is simultaneously
Corresponding first tracer I1With the second tracer I2。
Such as 30 structural block diagram of segmentation of lung parenchyma device that Fig. 4 is one embodiment of the invention.The segmentation of lung parenchyma device 30 packet
Image segmentation module 300 is included, which includes the first medical image acquisition unit 301, the life of the second medical image
At unit 302, skin line determination unit 303, cutting unit 304.Illustratively: the first medical image acquisition unit 301 can connect
Storage unit 103 is connect, for obtaining the first medical image of subject's scanned position, which includes several pieces
Tomographic image, and each slice image includes multiple pixels.In one embodiment, the first medical image is the first MR image.
Second medical image generation unit 302 is connect with the first medical image acquisition unit 301, for the first medicine
The sagittal plane of image is linearly enhanced, and to the first medical image and the first medical image respective pixel by linearly enhancing
Gray value be added, generate the second medical image further linearly enhanced it in the sagittal plane to the first medical image
Before, can also gamma correction processing be carried out to the first medical image.In one embodiment, the second medical image can be the first MR
Image is added the 2nd MR image obtained with the gray value by the first MR image respective pixel linearly enhanced.
Skin line determination unit 303 is connect, for determining the second medical image with the second medical image generation unit 302
Sagittal plane skin line, and sagittal plane skin line determine the second medical image on determine the second medical image cross-section musculus cutaneus
Skin line；
Cutting unit 304 is connect with skin line determination unit 303, for obtaining sagittal plane skin line and cross-section surface skin
The connected domain that line surrounds, and in connected domain obtain lung tissue segmentation result.It further, can also be according to the segmentation knot of lung tissue
Fruit is that each pixel distributes corresponding attenuation coefficient.It should be noted that in another embodiment, segmentation of lung parenchyma device 30 may be used also
It is more for imaging region to be divided into including skeletal tissue's cutting unit, musculature cutting unit, adipose tissue cutting unit etc.
Sub-regions, and each subregion only includes same tissue, realize imaging position respectively organize, the Accurate Segmentation of organ.
In some embodiments, image segmentation device 30 can realize the segmentation of multiple organ or tissues such as lung tissue, wherein
The dividing method of lung tissue includes: as shown in Figure 5
Step 510. obtains the first medical image of subject's scanned position, and the first medical image may include several lamellas
Image, and each slice image of the first medical image includes multiple pixels or voxel.Optionally, the first medical image can be
Three-dimensional magnetic resonance (MR) sequence image, computerized tomography (CT) image or positron emission tomography (PET) image etc., and
First medical image includes sagittal plane (sagittal plane), coronal-plane (coronal plane) and cross section
(transverse plane).If Fig. 6 is the first medicine that the first medical image acquisition of one embodiment of the invention unit 301 obtains
Image can be MR image, wherein Fig. 6 a is cross-sectional view strength；Fig. 6 b is coronal-plane view；Fig. 6 c is sagittal plane view.Three kinds
Different gray values respectively indicate different scanned positions in view, lung areas, background area and including bodies such as nasal cavity, oral cavities
The pixel of cavity region has low ash angle value；The pixel in the regions such as muscle, soft tissue has high gray value.Therefore, only according to ash
Angle value can not be by lung areas from medical image segmentation.
Step 520. linearly enhances the sagittal plane (gray value) of the first medical image, and to the first medical image and
Gray value by the first image respective pixel of grey level enhancement is added, and generates the second medical image.The arrow of first medical image
Shape face, which carries out grey level enhancement, may be expressed as: with formula
G (x, y)=T [f (x, y)] (formula 1)
Wherein, (x, y) indicates the position of the first medical image (sagittal plane) pixel, and x indicates the abscissa of pixel, y table
Show the ordinate of pixel；F (x, y) indicates that the first medical image converts the gray value for the pixel that preceding coordinate is (x, y)；G (x, y) table
Show that the first medical image transformation recoil is designated as the gray value of the pixel of (x, y)；T indicates certain mapping relations；First medical image
The actual range of gray scale is represented by [f before converting1,f2], the range required after the transformation of the first medical image is represented by [g1,
g2].It can be realized using following formula and gray scales are stretched or compressed, to achieve the effect that enhancing comparison.
In another embodiment, the sagittal plane of the first medical image, which carries out linear enhancing, can be used based on gradient of image and gray scale
Linear enhancement method, so that the profile of display foreground and background difference is become larger.It can refer to document Weickert,
J.1996.Anisotropic Diffusion in Image Processing.Ph.D.Thesis,Dept.of
Mathematics, University of Kaiserslautern, Germany, pp.42-43,80-82,107, to realize
The enhancing of target area and region intersection (skin line edge), optionally uses two Gaussian smoothing systems in one medical image
Number may be set to any number between 0.3-0.6.It in another embodiment, can be to the imaging organic region in the first medical image
Different greyscale transformation is respectively adopted with background area, and tonal range occupied by imaging organ is stretched, to background
It is compressed.In this embodiment, following steps can be executed in the second medical image generation unit 302: traversal first
Medical image all pixels obtain the maximum gradation value f of all pixelsmaxWith minimum gradation value fmin；Gray threshold is selected respectively
First medical image is split, obtains imaging organic region, background area and boundary in imaging organic region and background area
The transitional region of domain between the two；Stretching conversion is used to the gray level of imaging organic region, the gray level of transitional region is kept
It is constant, and compressed transform is used to the gray level of background area.In this particular embodiment, double gray scale thresholds may be selected in gray threshold
Value fth1And fth2, and fmin<fth1<fth2<fmax.When the gray value f of any pixel in the first medical image meets fmin<f<
fth1, then the pixel is classified as background area；When the gray value f of any pixel in the first medical image meets fth1<f<fth2,
Then the pixel is classified as transitional region；When the gray value f of any pixel in the first medical image meets fth2<f<fmax, then should
Pixel is classified as imaging organic region.Range [g that is corresponding, being required after the transformation of the first medical image1,g2] in setting two
A gray threshold gth1And gth2, and g1<gth1<gth2<g2.Imaging organic region, background area and transitional region can be adopted respectively
With following greyscale transformation mode:
By aforesaid operations, the edge for the skin line that can be enhanced.Further, by the first medical image of enhancing
The gray value corresponded at each pixel with original first medical image is added, and the second medical image can be obtained.It should be noted that
Before to the processing of the first medical image enhancement, can also gamma correction be carried out to the first medical image.More specifically, gamma correction mistake
Correction coefficient in journey may be selected between 1.05-1.45.
Step 530. determines the sagittal plane skin line of the second medical image, and the second medicine determined in sagittal plane skin line
The cross section skin line of the second medical image is determined on image.Illustratively, the sagittal plane skin line of the second medical image can lead to
The acquisition of skin line determination unit 303 is crossed, and executes step as shown in Figure 7 in the unit:
Step 710. carries out binary conversion treatment using given threshold to each pixel of the second medical image sagittal plane；
Step 720. counts the quantity of the included connected domain of each lamella of the second medical image sagittal plane；
Step 730. judges the quantity of the included connected domain of each lamella whether less than the first given threshold, if condition is full
Foot, then retain the connected domain, and execute step 750；Otherwise, step 740 is executed；In some embodiments, first setting
Threshold value may be provided at any integer value between 1-5.Only Retention area is maximum in the included connected domain of each lamella for step 740.
One or more connected domains, and execute step 750；Optionally, first given threshold is set as 2 in one embodiment,
At most retain two connected domains in each lamella by aforesaid operations, which is that area is maximum in all connected domains
The first connected domain and area only secondary first connected domain the second connected domain.
Step 750. determines the sagittal plane skin line of the second medical image according to the connected domain of reservation.
In an embodiment of the invention, in each pixel to the second medical image sagittal plane using setting pixel threshold
Before carrying out binary conversion treatment, three sagittal plane picture numbers can be converted by the three-dimensional cross-sectional image data of the second medical image
According to, and 30-60 may be selected in the range for setting pixel threshold.
Illustratively, the sagittal plane skin line of the second medical image is obtained by following process: obtaining slice image boundary
The point that upper pixel value is 1, and the connected domain of reservation is closed by the point for being 1 according to pixel value；The reservation connected domain of closure is carried out
The position that gray value in connected domain is 0 is assigned a value of 1 by filling, the boundary pixel set of connected domain is the second doctor after filling
Learn the sagittal plane skin line of image.In this particular embodiment, the closing course of the connected domain of reservation are as follows: for passing through step
730 connected domains retained are closed if there are intersection in the coboundary or lower boundary of connected domain and image on each lamella tomographic image
Lower boundary or coboundary.Illustrate for following closing of the frontier, boundary is located at the Nth row of slice image instantly, searches on N-1 row
Then the position (N-1, B) that first pixel value is 1 position (N-1, A) and the last one pixel value is 1 arranges Nth row from A
The pixel value arranged to B is assigned to 1, forms the connected domain of a closure, A, B respectively indicate the columns where pixel.
On the basis of obtaining the second medical image sagittal plane, the cross section of the second medical image can be determined based on this
Skin line.Illustratively, firstly, obtaining the second medical image that the sagittal plane skin line determines, which is
Three-dimensional sagittal plane image data；Then, each lamella in the second medical image cross section determined to sagittal plane skin line carries out
Two-dimentional corrosion treatment, Retention area are greater than the connected domain of the first setting area threshold；It is greater than first to the area of the reservation to set
The connected domain for determining area threshold carries out two-dimensional swelling processing, and according to two-dimensional swelling treated connected domain determines second doctor
Learn the cross section skin line of image.In above process, the first setting area threshold may be provided between 100-200.At one
In embodiment, the first setting area threshold may be set to 150.The second medical image cross section determining to sagittal plane skin line
Each slice image carries out the circle that two-dimentional etching operation available parameter radius is 3-6；Two-dimensional swelling handles available parameter
Radius is the circle of 2-4.
Step 540. obtains the connected domain that sagittal plane skin line and cross section skin line surround, and lung is obtained in connected domain
The segmentation result of tissue.Illustratively, following process can be performed in segmentation of lung parenchyma unit 304 as shown in Figure 8 and obtain lung group
The segmentation result knitted:
Step 810. (is surrounded) by sagittal plane skin line and cross section skin line in connected domain and obtains high gray scale in cross section
It is worth the boundary pixel point of connected domain, i.e., divides high gray value connected domain in connected domain cross section, obtain high gray value connected domain side
The collection of the point that pixel value is 1 in boundary, the point composition that pixel value is 1 is combined into boundary pixel point., illustratively, can be first according to arrow
The three-dimensional cross section greyscale image data that shape surface skin line and cross section skin line determine carries out Threshold segmentation, and (gray threshold is optional
Select 30-60), binary image is obtained, even the pixel that gray value of image is greater than gray threshold is 1, other pixels are 0.Then
The characteristics of being imaged according to lung tissue screens connected domain, in this embodiment, connected domain in statistics available each slice image
Number retains all connected domains if the number of connected domain is less equal than the second given threshold；If connected domain number is more
In the second given threshold, then retain the connected domain that quantity is equal to the second given threshold；The connected domain of reservation is further filled out
Processing is filled, and calculates the area of connected domain after filling, company of the connected domain area less than the second setting area threshold after removal filling
Logical domain, finally obtains high gray value region (connected domain).In some embodiments, the second given threshold may be configured as between 3-5
Any number.Optionally, the second given threshold may be set to 3, and the connected domain retained is I1、I2And I3.In one embodiment
In, enabling the connected domain of reservation is I1Area be S1, enabling the connected domain of reservation is I2Area be S2Enabling the connected domain retained is I3
Area be S3, the maximum connected domain I of area in connected domain to be removed4, and lead I4Area be S4, then should meet S1≥S4；S2
≥S4；S3≥S4.In another embodiment, the second setting area threshold may be provided at any number between 200-250.
Step 820. obtains cross section segmented image and coronal-plane segmented image in the connected domain that boundary pixel point surrounds,
Cross section segmented image or coronal-plane segmented image are low ash angle value region.In one embodiment, the picture that all pixels value is 1
The connected domain of reservation can be closed by vegetarian refreshments.There are low ash angle value regions in the closed communicating domain of above-mentioned reservation.It can be to high gray scale
The three-dimensional cross section greyscale image data of value determination carries out Threshold segmentation (gray threshold may be set to 30-60), obtains binaryzation
Image；The number of connected domain in each slice image is counted, if the number of connected domain is less than 2, retains all connected domains；
If connected domain number more than two, maximum 2 connected domains of Retention area are by the connected domain that aforesaid operations obtain
Cross section segmented image.Further, 3-D image region is determined for cross section segmented image, on its corresponding coronal-plane
The number of connected domain in each slice image is counted, if the number of connected domain is less than 2, retains all connected domains；If
Connected domain number more than two, then maximum 2 connected domains of Retention area, the connected domain obtained by aforesaid operations are as coronal
Face segmented image.
Step 830. obtains center line according to coronal-plane segmented image.In one embodiment, the acquisition process of center line
Can include: the connected domain area for calculating each lamella of coronal-plane segmented image determines the maximum lamella of connected domain area；It is being connected to
The initial row and termination row of connected domain are determined in the maximum lamella of domain area；According in the initial row of connected domain and termination row determination
Heart line.In an embodiment of the present invention, the gross area of connected domain in each slice image can be traversed for coronal-plane segmented image,
The area for comparing connected domain on each lamella can be obtained the maximum lamella S of connected domain area.The connected domain area is maximum
The bianry image of lamella S can be projected along Y-axis, find the initial row L of connected domain1With termination row L2, initial row L1That is lung tissue
The initial row in region, termination row L2The as termination row in lung tissue region, L1And L2Average value be line number where center line
L3。
Step 840. determines lung tissue pair in the position of cross section segmented image and coronal-plane segmented image according to center line
The connected domain answered.Illustratively, the step of connected domain corresponding according to center line acquisition lung tissue are as follows: in coronal-plane segmented image
Upper reservation center line connected domain of the row；Center line connected domain of the row is filled in cross section, obtains tissue pair
The connected domain answered.In a specific embodiment of the invention, it can retain in coronal-plane segmented image and pass through center line L3Be expert at (such as:
[L3- 2, L3+ 3] connected domain) or by image lower boundary being expert at (such as: [N-3, N])；To coronal-plane segmented image, conversion
Onto corresponding cross section, it is filled to connected domain is retained, obtains segmentation of lung parenchyma result.The lung obtained by the above process
Tissue segmentation result can be avoided influence of the background area to lung segmentation result, improve the accuracy of automatic segmentation, be applicable in
It is scanned in more beds, improves scanning speed.
In PET/MR multi-mode imaging system, it usually needs the information based on MR image carries out PET correction for attenuation: by MR
Image segmentation is divided into several regions, wherein each region separately includes multiple voxels for belonging to same tissue or organ,
Several regions can respectively correspond the voxel of the Different Organs such as skin, lung tissue, soft tissue and bone or tissue；According to priori
Information is that different attenuation coefficients is distributed in several regions of segmentation.
The dividing method of lung tissue of the present invention can accurately and efficiently realize the automatic segmentation of lung tissue in medical image.Such as
Fig. 9 a is the cross section skin line result schematic diagram that one embodiment of the invention obtains, and passes through the skin line (outside of lung areas
Contour line) background area and image tissue region of medical image can be distinguished, and include different gray scales inside image tissue region
It is worth region, and bone tissue and the air section of lung tissue show as high gray value simultaneously.Fig. 9 b is one embodiment of the invention acquisition
Segmentation of lung parenchyma result schematic diagram, above-mentioned skin line surround connected domain on the basis of, can accurately confirm the region A of lung tissue
And B, realize accurately identifying for the lung tissue air section that includes and bone tissue；In conjunction with gray scale of the lung tissue in coronal image
Distribution and location information avoid being applicable to more bed scannings using complicated priori matching template.
In other embodiments of the invention, above-mentioned Medical Image Processing module can be used for the correction for attenuation of PET imaging.Show
Example property, rebuilding module 400 can be according to the segmentation of lung parenchyma that image segmentation module 300 obtains as a result, correction PET data.It rebuilds
Module 400 can iterative approximation PET data, and according to after iterative approximation PET data generate PET image, wherein in PET data
The PET data is corrected using decay pattern during iterative approximation, and in the iterative reconstruction process of PET data in iteration update
State decay pattern.
Control module 500 can be centralization, such as data center；Be also possible to it is distributed, such as one distribution
Formula system.Control module 500 can be local, be also possible to long-range.In some embodiments, control module 500 can be with
Including central processing unit (Central Processing Unit, CPU), specialized application integrated circuit (Application
Specific Integrated Circuit, ASIC), dedicated instruction processor (Application Specific
Instruction Set Processor, ASIP), physical processor (Physics Processing Unit, PPU), number
Signal processor (Digital Processing Processor, DSP), field programmable gate array (Field-
Programmable Gate Array, FPGA), programmable logic device (Programmable Logic Device, PLD),
The combination of one or more of processor, microprocessor, controller, microcontroller etc..
Display module 600 may also display the height, weight, age, imaging position, MR scan module 100, PET of subject
The working condition of scan module 200, the magnetic resonance image that position is imaged or PET image etc..The type of display module 600 can be
In cathode-ray tube (CRT) display, liquid crystal display (LCD), organic light emitting display (OLED), plasma display etc.
One or more of combinations.
It should be noted that each module of magic magiscan of the invention or unit can connect between each other, the company
It connects and can be wireless network connection or finite element network connection.Wherein, cable network may include utilizing metallic cable, mixing electricity
One or more combined modes such as cable, one or more interfaces.Wireless network may include utilizing bluetooth, regional area networks
(LAN), one or more groups of wide local area network (WAN), near source field communication (Near Field Communication, NFC) etc.
The mode of conjunction.
Above for the description of magic magiscan, only for convenience of description, the application can not be limited in and be lifted
Within scope of embodiments.It is appreciated that for those skilled in the art, after the principle for understanding the system, Ke Neng
In the case where without departing substantially from this principle, any combination is carried out to modules, or constitute subsystem and connect with other modules, it is right
Implement the various modifications and variations of the above method and systematic difference field in form and details.
In one embodiment, the process of magic magiscan processing multi-modality medical image are as follows: scanned using MR
Module 100 scans subject's organic region, obtains the MR image of corresponding organic region, and is stored in storage unit 103；Using
PET scan module 200 scans subject's organic region, obtains the PET data of corresponding organic region, and is stored in memory 204；
Image segmentation module 300 obtains MR image from storage unit 103, obtains sagittal plane skin line and cross-section surface skin from MR image
Line, and the corresponding connected domain of lung tissue is determined according to skin line, and, it is determined in connected domain according to the structure feature of lung tissue
The segmentation result of lung tissue；Module 400 is rebuild to connect with segmentation of lung parenchyma unit 304, memory 204, it can be according to lung tissue point
The segmentation result for cutting the acquisition of unit 304 is that each pixel (or voxel) distributes respective attenuation coefficient, generates the first decay pattern；According to
One decay pattern rebuilds PET data, obtains the first PET image；The first decay pattern is updated according to the first PET image, and generates second
Decay pattern；The first PET image is rebuild based on the second decay pattern, obtains the second PET image；It repeats the above process and knows production
Raw final target decay pattern (decay pattern estimation) and final goal PET image (PET image estimation).
Further, display module 600 can show the multi-modality images of PET image and MR image co-registration, and image melts
Light stream field method, the method for registering based on characteristic point, the method for registering based on appearance profile or based on gray value etc. can be used in conjunction
Method for registering.
It should be noted that through the above description of the embodiments, those skilled in the art can be understood that
It can be realized to some or all of the present invention by software and in conjunction with required general hardware platform.Based on this understanding,
Substantially the part that contributes to existing technology can embody technical solution of the present invention in the form of software products in other words
Out, which may include the one or more machine readable medias for being stored thereon with machine-executable instruction,
These instructions may make this when being executed by one or more machines such as computer, computer network or other electronic systems
One or more machine embodiment according to the present invention execute operation.
Although the present invention is disclosed as above with preferred embodiment, however, it is not to limit the invention, any this field skill
Art personnel, without departing from the spirit and scope of the present invention, when can make a little modification and perfect therefore of the invention protection model
It encloses to work as and subject to the definition of the claims.
Claims (9)
1. a kind of segmentation of lung parenchyma method, includes the following steps:
The first medical image of subject target area is obtained, first medical image includes several slice images, described
Each slice image includes multiple pixels；
The sagittal plane of first medical image is linearly enhanced, and to first medical image and by linear enhancing
The first image respective pixel gray value be added, generate the second medical image；
Determine the sagittal plane skin line of second medical image, and the second medical image determined in the sagittal plane skin line
The cross section skin line of upper determination second medical image；
The connected domain that the sagittal plane skin line and cross section skin line surround is obtained, and obtains lung tissue in the connected domain
Segmentation result；
The segmentation result of the lung tissue is obtained by following process: obtaining high gray value connection in the cross section of the connected domain
The boundary pixel point in domain；Cross section segmented image and coronal-plane segmentation figure are obtained in the connected domain that the boundary pixel point surrounds
Picture, the cross section segmented image or the coronal-plane segmented image are low ash angle value region；According to the coronal-plane segmentation figure
As obtaining center line；It is determined according to the center line in the position of the cross section segmented image and the coronal-plane segmented image
The corresponding connected domain of lung tissue.
2. segmentation of lung parenchyma method according to claim 1, which is characterized in that first medical image be CT image or
MR image.
3. segmentation of lung parenchyma method according to claim 1, which is characterized in that the sagittal musculus cutaneus of second medical image
Skin line is obtained by following process:
Binary conversion treatment is carried out using given threshold to each pixel of the second medical image sagittal plane；
Count the quantity of the included connected domain of each lamella of the second medical image sagittal plane after binary conversion treatment；
The quantity of the included connected domain of each lamella is judged whether less than the first given threshold, if the conditions are met, then described in reservation
Connected domain；Otherwise, the only maximum one or more connected domains of Retention area in included connected domain；
The sagittal plane skin line of second medical image is determined according to the connected domain of the reservation.
4. segmentation of lung parenchyma method according to claim 3, which is characterized in that further include obtaining picture on slice image boundary
The point that element value is 1, and the connected domain of reservation is closed by the point for being 1 according to the pixel value.
5. segmentation of lung parenchyma method according to claim 1, which is characterized in that the cross-section musculus cutaneus of second medical image
Skin line is determined by following process:
Obtain the second medical image that the sagittal plane skin line determines；
Two-dimentional corrosion treatment is carried out to each lamella in the second medical image cross section that the sagittal plane skin line determines, is retained
Area is greater than the connected domain of the first setting area threshold；
The connected domain for being greater than the first setting area threshold to the area of the reservation carries out two-dimensional swelling processing, and swollen according to two dimension
The cross section skin line of swollen treated connected domain determines second medical image.
6. segmentation of lung parenchyma method according to claim 1, which is characterized in that the center line is obtained by following process
It takes:
The connected domain area for calculating each lamella of coronal-plane segmented image, determines the maximum lamella of connected domain area；
The initial row and termination row of connected domain are determined in the maximum lamella of connected domain area；
Center line is determined according to the initial row of the connected domain and termination row.
7. segmentation of lung parenchyma method according to claim 1, which is characterized in that according to the center line in the cross section
The position of segmented image and the coronal-plane segmented image determines the specific steps of the corresponding connected domain of lung tissue are as follows:
Retain center line connected domain of the row in the coronal-plane segmented image；
Center line connected domain of the row is filled in the cross section segmented image, it is corresponding to obtain lung tissue
Connected domain.
8. a kind of segmentation of lung parenchyma device, including image segmentation module, the image segmentation module include:
First medical image acquisition unit, for receiving the first medical image of subject target area, the first medicine figure
As including several slice images, each slice image includes multiple pixels；
Second medical image generation unit is linearly enhanced for the sagittal plane to first medical image, and to described
First medical image is added with the gray value by the first medical image respective pixel linearly enhanced, generates the second medicine figure
Picture；
Skin line determination unit, for determining the sagittal plane skin line of second medical image, and in the sagittal surface skin
The cross section skin line of second medical image is determined on the second medical image that line determines；
Cutting unit, the connected domain surrounded for obtaining the sagittal plane skin line and cross section skin line, and in the connection
The segmentation result of lung tissue is obtained in domain, the segmentation result of the lung tissue is obtained by following process:
The boundary pixel point of high gray value connected domain is obtained in the cross section of the connected domain；It is surrounded in the boundary pixel point
Cross section segmented image and coronal-plane segmented image, the cross section segmented image or coronal-plane segmentation are obtained in connected domain
Image is low ash angle value region；Center line is obtained according to the coronal-plane segmented image；According to the center line described cross-section
The position of face segmented image and the coronal-plane segmented image determines the corresponding connected domain of lung tissue.
9. a kind of magic magiscan, which is characterized in that the magic magiscan includes as claimed in claim 8
Segmentation of lung parenchyma device, further includes:
MR scan module for scanning subject target area, and obtains corresponding first medical image in the target area；
PET scan module for scanning the target area, and acquires the corresponding PET data in target Europe region；
Rebuild module, for obtain segmentation of lung parenchyma that the segmentation of lung parenchyma device obtains as a result, and be the segmentation lung
Each pixel distribution respective attenuation coefficient is organized, decay pattern, and the PET data according to the decay pattern iterative approximation are generated
Generate PET image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610912654.0A CN106558045B (en) | 2016-10-20 | 2016-10-20 | A kind of segmentation of lung parenchyma method, apparatus, magic magiscan |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610912654.0A CN106558045B (en) | 2016-10-20 | 2016-10-20 | A kind of segmentation of lung parenchyma method, apparatus, magic magiscan |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106558045A CN106558045A (en) | 2017-04-05 |
| CN106558045B true CN106558045B (en) | 2019-07-19 |
Family
ID=58443306
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610912654.0A Active CN106558045B (en) | 2016-10-20 | 2016-10-20 | A kind of segmentation of lung parenchyma method, apparatus, magic magiscan |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106558045B (en) |
Families Citing this family (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113035327A (en) | 2016-10-27 | 2021-06-25 | 普罗热尼奇制药公司 | Network, decision support system and related Graphical User Interface (GUI) application for medical image analysis |
| CN107260137A (en) * | 2017-07-31 | 2017-10-20 | 济南显微智能科技有限公司 | It is a kind of to detect ICG and double Tracing detection device and methods of methylenum careuleum simultaneously |
| CN107563998B (en) * | 2017-08-30 | 2020-02-11 | 上海联影医疗科技有限公司 | Method for processing heart image in medical image |
| EP3738097A2 (en) * | 2018-01-08 | 2020-11-18 | Progenics Pharmaceuticals, Inc. | Systems and methods for rapid neural network-based image segmentation and radiopharmaceutical uptake determination |
| JPWO2019220825A1 (en) * | 2018-05-16 | 2021-06-17 | パナソニック株式会社 | Chest X-ray image gradation conversion method, image gradation conversion program, image gradation conversion device, server device, and conversion method |
| CN110555860B (en) * | 2018-06-04 | 2023-11-14 | 青岛海信医疗设备股份有限公司 | Method for labeling rib areas in medical image, electronic equipment and storage medium |
| CN109512449A (en) * | 2018-09-26 | 2019-03-26 | 沈阳东软医疗系统有限公司 | Physiological parameter output method and device |
| KR101981202B1 (en) * | 2018-12-11 | 2019-05-22 | 메디컬아이피 주식회사 | Method and apparatus for reconstructing medical image |
| CN113272859A (en) | 2019-01-07 | 2021-08-17 | 西尼诊断公司 | System and method for platform neutral whole body image segmentation |
| EP3959684A1 (en) | 2019-04-24 | 2022-03-02 | Progenics Pharmaceuticals, Inc. | Systems and methods for automated and interactive analysis of bone scan images for detection of metastases |
| CN110363744B (en) * | 2019-06-10 | 2021-08-06 | 深圳市智影医疗科技有限公司 | Lung age detection method and equipment |
| US11900597B2 (en) | 2019-09-27 | 2024-02-13 | Progenics Pharmaceuticals, Inc. | Systems and methods for artificial intelligence-based image analysis for cancer assessment |
| CN110689550B (en) * | 2019-10-12 | 2022-06-21 | 嘉应学院 | High-efficiency automatic screening system and method for CT (computed tomography) images of sagittal plane of lumbar vertebra |
| CN111128317A (en) * | 2019-11-20 | 2020-05-08 | 中国辐射防护研究院 | Ionizing radiation tissue equivalent material formula design method and system |
| CN112884792B (en) * | 2021-02-02 | 2022-10-25 | 青岛海信医疗设备股份有限公司 | Lung image segmentation method and device, electronic equipment and storage medium |
| CN114943856B (en) * | 2021-04-12 | 2024-04-26 | 四川省肿瘤医院 | Lung nodule region identification method, labeling method and identification system |
| CN115937163B (en) * | 2022-12-19 | 2023-07-28 | 西北民族大学 | Method and system for extracting target region of SPECT (single photon emission computed tomography) lung perfusion imaging |
| CN116205920B (en) * | 2023-05-05 | 2023-07-18 | 天津医科大学总医院 | Method and system for generating key region detection model based on lung scanning data |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7309867B2 (en) * | 2003-04-18 | 2007-12-18 | Medispectra, Inc. | Methods and apparatus for characterization of tissue samples |
| JP2009525780A (en) * | 2006-02-03 | 2009-07-16 | コーニンクレッカ フィリップス エレクトロニクス エヌ ヴィ | Clarifying heterogeneous objects when creating CT-based attenuation maps |
| US9597041B2 (en) * | 2007-03-30 | 2017-03-21 | General Electric Company | Sequential image acquisition with updating method and system |
| US20100158332A1 (en) * | 2008-12-22 | 2010-06-24 | Dan Rico | Method and system of automated detection of lesions in medical images |
| JP5804340B2 (en) * | 2010-06-10 | 2015-11-04 | 株式会社島津製作所 | Radiation image region extraction device, radiation image region extraction program, radiation imaging device, and radiation image region extraction method |
| JP6603245B2 (en) * | 2014-07-02 | 2019-11-06 | コヴィディエン リミテッド パートナーシップ | System and method for lung segmentation |
| CN104992445B (en) * | 2015-07-20 | 2017-10-20 | 河北大学 | A kind of automatic division method of CT images pulmonary parenchyma |
-
2016
- 2016-10-20 CN CN201610912654.0A patent/CN106558045B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN106558045A (en) | 2017-04-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106558045B (en) | A kind of segmentation of lung parenchyma method, apparatus, magic magiscan | |
| CN106600609B (en) | Spine segmentation method and system in medical image | |
| CN107644421B (en) | Medical image segmentation method and system | |
| CN109961834B (en) | Image diagnosis report generation method and device | |
| CN105263407B (en) | Use the detection to bone tissue of magnetic resonance imaging | |
| US9204817B2 (en) | Attenuation correction in positron emission tomography using magnetic resonance imaging | |
| EP2252204B1 (en) | Ct surrogate by auto-segmentation of magnetic resonance images | |
| Kartmann et al. | Integrated PET/MR imaging: automatic attenuation correction of flexible RF coils | |
| CN102525466B (en) | Image processing apparatus and MR imaging apparatus | |
| US8218848B2 (en) | System and method for the generation of attenuation correction maps from MR images | |
| WO2017048856A1 (en) | Simultaneous ct-mri image reconstruction | |
| CN103445801B (en) | For performing the method and system of correction for attenuation | |
| US8867814B2 (en) | Methods and systems for generating a positron emission tomography attenuation correction map | |
| EP2245592B1 (en) | Image registration alignment metric | |
| US20180356484A1 (en) | Removal of image artifacts in sense-mri | |
| CN107292889B (en) | Tumor segmentation method, system and readable medium | |
| Eldib et al. | Markerless attenuation correction for carotid MRI surface receiver coils in combined PET/MR imaging | |
| WO2020033566A1 (en) | Neural networks for volumetric segmentation and parcellated surface representations | |
| CN116630342A (en) | Abdominal MRI image segmentation system, method, electronic device, and storage medium | |
| Rybak et al. | Measurement of the upper respiratory tract aerated space volume using the results of computed tomography | |
| Könik et al. | Digital anthropomorphic phantoms of non-rigid human respiratory and voluntary body motions: A tool-set for investigating motion correction in 3D reconstruction | |
| Frohwein et al. | Correction of MRI‐induced geometric distortions in whole‐body small animal PET‐MRI | |
| US20240046535A1 (en) | System and method for pet-ct image artifact evaluation and correction | |
| JP2019505253A (en) | Removal of image artifacts in SENSE imaging | |
| Haukipuro | 3D geometric correction in 2D multi-slice acquired MR images used in radiation therapy planning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP01 | Change in the name or title of a patent holder | ||
| CP01 | Change in the name or title of a patent holder |
Address after: 201807 Shanghai city Jiading District Industrial Zone Jiading Road No. 2258 Patentee after: Shanghai Lianying Medical Technology Co., Ltd Address before: 201807 Shanghai city Jiading District Industrial Zone Jiading Road No. 2258 Patentee before: SHANGHAI UNITED IMAGING HEALTHCARE Co.,Ltd. |