CN114334130B - Brain symmetry-based PET molecular image computer-aided diagnosis system - Google Patents
Brain symmetry-based PET molecular image computer-aided diagnosis system Download PDFInfo
- Publication number
- CN114334130B CN114334130B CN202111605249.1A CN202111605249A CN114334130B CN 114334130 B CN114334130 B CN 114334130B CN 202111605249 A CN202111605249 A CN 202111605249A CN 114334130 B CN114334130 B CN 114334130B
- Authority
- CN
- China
- Prior art keywords
- brain
- image
- symmetrical
- template
- images
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Magnetic Resonance Imaging Apparatus (AREA)
- Nuclear Medicine (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a PET molecular image computer-aided diagnosis system based on brain symmetry, which comprises: the system comprises an image acquisition and preprocessing module, an asymmetric brain template construction module, a symmetric brain template construction module, a symmetry analysis module and an auxiliary diagnosis module. The system constructs an asymmetric brain template according to the normal brain PET molecular image, and further constructs a symmetric brain template; registering the individual images to a symmetrical brain template space to obtain priori knowledge of a symmetrical structure, and introducing individual image information by using a symmetrical differential stratospheric algorithm to obtain a brain structure symmetrical relation of a voxel level; and finally, comprehensively utilizing the metabolism or receptor information of the focus area and the symmetrical brain area thereof to assist in improving the diagnosis capability of the PET molecular image.
Description
Technical Field
The invention belongs to the technical field of computer-aided diagnosis of medical images, and particularly relates to a PET molecular image computer-aided diagnosis system and method based on brain symmetry.
Background
Positron emission computed tomography (PET) is a molecular imaging technique capable of visualizing molecular metabolism, receptor and neurotransmitter activity in vivoHas been widely used for diagnosis, differential diagnosis, disease assessment, efficacy assessment, etc. of various neurological diseases. For example, the number of the cells to be processed, 18 fluoro-fluoro deoxyglucose 18 F-FDG) PET can detect brain glucose metabolism abnormality caused by diseases, and has become a conventional imaging examination means for glioma diagnosis, refractory type preoperative epilepsy evaluation and the like; 11 c-2 beta-methyl ester-3 beta- (4-fluorophenyl) tropane 11 C-CFT) PET can visualize the distribution of dopamine transporter in brain, and has important value in the typing of Parkinson's disease.
Clinical analysis of PET molecular images mainly relies on visual assessment of physicians, but visual assessment of three-dimensional PET images is very time-consuming, and severely depends on clinical knowledge and working experience of physicians, and is highly subjective and poor in repeatability between different physician diagnosis results. Therefore, there is a need to develop a computer-aided diagnosis system to aid physicians in image diagnosis, to reduce excessive reliance on physician experience, and to improve accuracy and repeatability of examination results. The current computer-aided diagnosis system for PET molecular images comprises: 1) The method comprises the steps of (1) analyzing a region of interest (ROI), manually sketching or mapping by a doctor to obtain a Standard Uptake Value (SUV) of a brain symmetrical brain region, and diagnosing according to an Asymmetry Index (AI) of the SUV, wherein the method has high stability but poor sensitivity; 2) Based on Voxel-based statistical analysis (Voxel-based) the method diagnoses lesions by Voxel-by-Voxel contrast of SUVs or differences in standard uptake rates (SUVR) of patients and healthy people, which is highly sensitive but ignores brain uptake symmetry information of significant value.
In view of the foregoing, there is a need for developing a high-sensitivity, voxel-level PET molecular image computer-aided diagnosis system that can effectively utilize the imaging characteristics of a lesion and its contralateral brain region to improve the diagnostic capabilities of PET molecular images.
Disclosure of Invention
The invention provides a brain symmetry-based PET molecular image computer-aided diagnosis system, which realizes brain symmetry analysis on metabolism or receptor activity at voxel level and provides an effective tool for segmentation, positioning, diagnosis and the like of PET molecular images.
The aim of the invention is realized by the following technical scheme: a PET molecular image computer-aided diagnosis system based on brain symmetry comprises an image acquisition and preprocessing module, an asymmetric brain template construction module, a symmetric brain template construction module, a symmetry analysis module and an aided diagnosis module;
the image acquisition and preprocessing module is used for extracting brain areas of PET molecular images with the same body position, and calculating SUVR values by taking SUV average values of the selected brain areas as references;
the asymmetric brain template construction module is used for constructing an asymmetric brain template by using a registration-average iteration optimization method based on the normal brain PET molecular image processed by the image acquisition and preprocessing module;
the symmetrical brain template construction module utilizes affine registration to enable the vector surface in the asymmetrical brain template to be positioned in the center of the image to obtain a centralized asymmetrical brain template, then utilizes a symmetrical differential homoembryo method to register the centralized asymmetrical brain template and mirror images thereof, acquires median images in two registration processes, averages all corresponding pixel points of the two median images to generate an average image, and obtains the symmetrical brain template;
the symmetry analysis module is used for deforming the individual PET molecular images into a symmetrical brain template image space, registering the deformed images and mirror images thereof by using a symmetrical differential stratospheric method, and obtaining individual symmetrical images in the symmetrical brain template image space;
the auxiliary diagnosis module adopts two optional modes to realize auxiliary diagnosis:
a. carrying out statistical analysis based on voxels by utilizing individual symmetrical images in a symmetrical brain template image space, carrying out statistical comparison on the individual images and normal brain symmetrical images at each pixel point in the symmetrical brain template image space, and extracting pixel points with statistically significant spatial connection, namely brain regions with abnormal molecular metabolism or receptor activity;
b. based on individual symmetrical images in the symmetrical brain template image space, the symmetrical images in the individual space are obtained by inverse transformation and synchronously input into a machine learning model for auxiliary diagnosis.
Further, the image acquisition and preprocessing module adopts the extraction method of the same body position: the method comprises the steps of pre-defining a larger area for clipping through the area where the brain to be tested is in the vicinity in the scanning process, and clipping a specific brain area through a rigid registration technology to remove unnecessary tissue structures and backgrounds.
Further, the brain region selected in the image acquisition and preprocessing module comprises the whole brain, cerebellum grey matter or a brain bridge.
Further, the asymmetric brain template construction module is based on the normal brain PET molecular image (J) processed by the image acquisition and preprocessing module 1 ,J 2 ,…J n )，J n For the nth PET molecular image, two objective functions are synchronously optimized by using a registration-average iteration optimization method, and an asymmetric brain template phi is constructed:
wherein θ i For the ith PET molecular image J i Registering to an approximate templateIs a deformation of (1) in which the template is approximated>The approximate solution of the symmetrical brain template obtained finally is an average image of the images deformed to the asymmetrical brain template space in the previous iteration process; the asymmetric brain template construction process comprises two steps of registration and averaging; in the t-th iteration, "registration" registers normal brain PET molecular images to approximate brain template +.>"average" optimizes the approximate brain templates: the specific process is as follows: extracting deformation of the ith iteration->Non-rigid part in the inverse deformation of (2), averaging the non-rigid part as residual deformation R (t) The method comprises the steps of carrying out a first treatment on the surface of the Using residual deformation R (t) Correction of deformation-> By means of the corrected deformation->Deforming the normal brain PET molecular image into an asymmetric brain template space; the deformed normal brain PET molecular images are averaged to be used as an approximate brain template of the t+1st iteration: />Wherein->For using the corrected deformation +.>For image J i And (3) a result of the deformation.
Further, wherein the PET molecular image comprises 18 F-FDG PET 11 CFT-PET。
Further, the machine learning model used by the auxiliary diagnostic module includes a convolutional neural network, logistic regression, support vector machine, or random forest.
The invention has the beneficial effects that:
1. constructing PET molecular image brain templates of people, providing standard distribution of molecular metabolism and receptor activity, and providing standard public space for registration and the like;
2. the symmetrical brain template is explicitly constructed by utilizing a differential stratospheric algorithm, so that the detail information of the anatomical structure of the high-change area of the cortex is greatly reserved;
3. the brain symmetry analysis of the molecular metabolism and/or the receptor activity at the voxel level is realized, the analysis utilizes the priori knowledge of a high-precision symmetrical brain template, and the differential stratosphere algorithm is used for further analyzing the symmetrical structure of the individual image, so that the dislocation area of the tissue structure is greatly reduced;
4. the symmetrical information of the metabolism and the receptor is introduced into auxiliary diagnosis, so that the visual evaluation process of doctors is simulated, and the sensitivity of small focus detection can be improved.
Drawings
FIG. 1 is a block diagram of a brain symmetry based PET molecular image computer aided diagnosis system according to one embodiment of the present invention;
FIG. 2 is a block diagram of the construction of an asymmetric brain template in accordance with one embodiment of the present invention;
FIG. 3 is a block diagram of the construction of a symmetrical brain template according to one embodiment of the present invention.
Detailed Description
The invention will be described in further detail with reference to the drawings and the specific examples.
As shown in fig. 1, a brain symmetry-based PET molecular image computer-aided diagnosis system according to an embodiment of the present invention includes the following modules:
1. and the image acquisition and preprocessing module is used for: the three-dimensional PET molecular image of the tested is acquired by using PET/CT or PET/MR, the tested adopts the same body position in the acquisition process, and the PET molecular image comprises 18 F-FDG PET 11 CFT-PET. The acquired original DICOM formatted PET images are converted to easy-to-process NIfTI formatted images and the SUV values are calculated voxel-by-voxel. For the whole body scanned PET molecular image, the upper 1/5-1/4 image is intercepted to obtain the head PET image. Carrying out rigid registration (3 rotation angle parameters and 3 translation parameters) of 6 parameters on the head PET image and the brain template, deforming the brain template image into an individual PET image, obtaining the image range of the deformed PET brain template, and carrying out image processing on the individual head PET image according to the rangeAnd cutting out the lines, thereby obtaining a standardized brain PET image. And registering and deforming the standardized brain PET image to a brain template by affine transformation and nonlinear transformation, and selecting a whole brain SUV mean value, a cerebellum gray SUV mean value or a brain bridge SUV mean value as a reference to calculate a SUVR value according to requirements.
2. Asymmetric brain template construction module: as shown in fig. 2, the image acquisition and preprocessing module processes the normal brain PET image (J 1 ,J 2 ,…J n )，J n For an nth PET molecular image, synchronously optimizing two objective functions by using a registration-average iteration optimization method, and constructing an asymmetric brain template phi;
wherein θ i For the ith PET molecular image J i Registering to an approximate templateIs a deformation of (1) in which the template is approximated>The approximate solution of the symmetrical brain template obtained finally is obtained by the previous iteration process. The asymmetric brain template construction process includes two steps, registration and averaging. In the t-th iteration, "registration" registers normal brain PET molecular images to approximate brain template +.>The "average" optimizes the approximate brain template as follows:
(2.1) extracting the deformation of the ith iterationNon-rigid part in the inverse deformation of (2) and taking the average as residual deformation R (t) ；
(2.2) use of residual deformation R (t) Correcting deformation
(2.3) utilizing the corrected deformationDeforming the normal brain PET image to a template space;
(2.4) averaging the deformed normal brain PET images as an approximate brain template for the next iteration:wherein->For using the corrected deformation +.>For image J i And (3) a result of the deformation.
The construction process is iterated until convergenceAnd->The average squared distance of (2) is less than a preset threshold. A "coarse-fine" build procedure may be chosen, i.e. affine registration and non-linear registration are used sequentially. In affine registration, the deformation is corrected using matrix multiplication; in non-linear registration, vector field multiplication is used to correct the deformation: d (D) 1 ·D 2 (v)＝D 1 (v)+D 2 (v+D 1 (v) And D) wherein 1 And D 2 Is a deformation field, v isAnd (5) space coordinates.
3. Symmetrical brain template construction module: as shown in fig. 3, a symmetrical brain template is constructed from an asymmetrical brain template by using a symmetrical differential homoembryo registration algorithm, and the construction process comprises:
(3.1) centering: the asymmetric brain templates are mirror-inverted left and right, affine transformation is used for registering the asymmetric brain templates and mirror images thereof, and the affine transformation matrix is used for obtaining the square root deformed asymmetric brain templates, so that the vector surface is positioned in the center of the template space;
(3.2) symmetrization: registering the centered asymmetric brain template and the mirror image thereof by using a symmetrical differential homoembryo algorithm after left and right mirror image overturning centering, extracting a fixed image-median image transformation and a moving image-median image transformation, and respectively deforming the centered asymmetric brain template and the mirror image thereof to serve as two median images of the differential homoembryo algorithm;
and (3.3) averaging all corresponding pixels of the two median images to generate an average image as a symmetrical brain template.
4. Symmetry analysis module: and obtaining the symmetrical relation of the pixel level by using a differential embryo algorithm.
(4.1) registering the individual PET molecular images to a symmetric brain template using affine transformation and nonlinear transformation;
(4.2) mirror-turning left and right, carrying out symmetrical differential homoembryo registration, extracting fixed image-median image transformation, and applying the image to the registered image to obtain a symmetrical individual image, wherein voxels at the symmetrical position of the median vector surface are symmetrical structures of a pair of identical brain areas;
(4.3) (optionally) registering the symmetrical individual images with the symmetrical brain templates, wherein the bilateral symmetry of the deformation field is ensured in the registering process, and the individual symmetrical images of the symmetrical brain template space can be obtained.
5. An auxiliary diagnosis module: including two alternative implementations.
a. Statistical comparison analysis based on voxels: in the symmetrical brain template space, carrying out statistical comparison on the tested symmetrical image and the normal brain symmetrical image by voxels, such as a general linear model; multiple comparison corrections, such as FDR corrections, may be performed; and extracting voxels with obvious statistical test, namely brain regions with abnormal molecular metabolism or receptor activity.
b. Based on individual symmetrical images in the symmetrical brain template image space, the symmetrical images in the individual space are obtained by utilizing inverse transformation, the individual symmetrical images and left and right mirror images thereof are input into a machine learning model such as a convolutional neural network, logistic regression, a support vector machine or random forest for auxiliary diagnosis, such as the convolutional neural network, and segmentation, diagnosis, differential diagnosis and the like can be performed.
The present invention is not limited to the above-described preferred embodiments. Any person who can obtain other various imaging computer-aided diagnosis systems based on brain symmetry under the teaching of the present invention shall fall within the scope of the present invention.
Claims (5)
1. The PET molecular image computer-aided diagnosis system based on brain symmetry is characterized by comprising an image acquisition and preprocessing module, an asymmetric brain template construction module, a symmetric brain template construction module, a symmetry analysis module and an aided diagnosis module;
the image acquisition and preprocessing module is used for extracting brain areas of PET molecular images with the same body position, and calculating SUVR values by taking SUV average values of the selected brain areas as references;
the asymmetric brain template construction module is based on the normal brain PET molecular image (J) processed by the image acquisition and preprocessing module 1 ,J 2 ,…J n )，J n For the nth PET molecular image, two objective functions are synchronously optimized by using a registration-average iteration optimization method, and an asymmetric brain template phi is constructed:
wherein θ i For the ith PET molecular image J i Registering to an approximate templateIs a deformation of (1) in which the template is approximated>The approximate solution of the symmetrical brain template obtained finally is an average image of the images deformed to the asymmetrical brain template space in the previous iteration process; the asymmetric brain template construction process comprises two steps of registration and averaging; in the t-th iteration, "registration" registers normal brain PET molecular images to approximate brain template +.>"average" optimizes the approximate brain templates: the specific process is as follows: extracting deformation of the ith iteration->Non-rigid part in the inverse deformation of (2), averaging the non-rigid part as residual deformation R (t) The method comprises the steps of carrying out a first treatment on the surface of the Using residual deformation R (t) Correction of deformation->By means of the corrected deformation->Deforming the normal brain PET molecular image into an asymmetric brain template space; the deformed normal brain PET molecular images are averaged to be used as an approximate brain template of the t+1st iteration: />Wherein->For using the corrected deformation +.>For image J i And (3) a result of the deformation.
The symmetrical brain template construction module utilizes affine registration to enable the vector surface in the asymmetrical brain template to be positioned in the center of the image to obtain a centralized asymmetrical brain template, then utilizes a symmetrical differential homoembryo method to register the centralized asymmetrical brain template and mirror images thereof, acquires median images in two registration processes, averages all corresponding pixel points of the two median images to generate an average image, and obtains the symmetrical brain template;
the symmetry analysis module is used for deforming the individual PET molecular images into a symmetrical brain template image space, registering the deformed images and mirror images thereof by using a symmetrical differential stratospheric method, and obtaining individual symmetrical images in the symmetrical brain template image space;
the auxiliary diagnosis module adopts two optional modes to realize auxiliary diagnosis:
a. carrying out statistical analysis based on voxels by utilizing individual symmetrical images in a symmetrical brain template image space, carrying out statistical comparison on the individual images and normal brain symmetrical images at each pixel point in the symmetrical brain template image space, and extracting pixel points with statistically significant spatial connection, namely brain regions with abnormal molecular metabolism or receptor activity;
b. based on individual symmetrical images in the symmetrical brain template image space, the symmetrical images in the individual space are obtained by inverse transformation and synchronously input into a machine learning model for auxiliary diagnosis.
2. The brain symmetry-based PET molecular image computer-aided diagnosis system of claim 1, wherein the image acquisition and preprocessing module adopts the same body position extraction method: the method comprises the steps of pre-defining a larger area for clipping through the area where the brain to be tested is in the vicinity in the scanning process, and clipping a specific brain area through a rigid registration technology to remove unnecessary tissue structures and backgrounds.
3. The brain symmetry-based PET molecular image computer-aided diagnosis system of claim 1, wherein the selected brain regions in said image acquisition and preprocessing module include whole brain, cerebellum grey matter or brain bridge.
4. The brain symmetry-based PET molecular image computer-aided diagnosis system of claim 1, wherein the PET molecular image comprises 18 F-FDG PET 11 CFT-PET。
5. The brain symmetry-based PET molecular imaging computer-aided diagnosis system of claim 1, wherein the machine learning model used by the aided diagnosis module comprises a convolutional neural network, logistic regression, support vector machine or random forest.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111605249.1A CN114334130B (en) | 2021-12-25 | 2021-12-25 | Brain symmetry-based PET molecular image computer-aided diagnosis system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111605249.1A CN114334130B (en) | 2021-12-25 | 2021-12-25 | Brain symmetry-based PET molecular image computer-aided diagnosis system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114334130A CN114334130A (en) | 2022-04-12 |
| CN114334130B true CN114334130B (en) | 2023-08-22 |
Family
ID=81013960
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111605249.1A Active CN114334130B (en) | 2021-12-25 | 2021-12-25 | Brain symmetry-based PET molecular image computer-aided diagnosis system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114334130B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115005802B (en) * | 2022-07-21 | 2023-04-28 | 首都医科大学宣武医院 | Method, system and device for positioning initial part of brain network disease onset |
| CN115312181B (en) * | 2022-07-22 | 2023-05-09 | 首都医科大学宣武医院 | Prognosis judging method, system and device for brain network disease operation |
| CN116245833A (en) * | 2023-02-13 | 2023-06-09 | 复旦大学 | Prediction method for brain multi-mode PET image SUVR |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110335665A (en) * | 2019-04-01 | 2019-10-15 | 佛山原子医疗设备有限公司 | It is a kind of applied to medical image auxiliary diagnosis analysis to scheme to search drawing method and system |
| CN111009324A (en) * | 2019-11-25 | 2020-04-14 | 东北大学 | Mild cognitive impairment auxiliary diagnosis system and method based on brain network multi-feature analysis |
| KR102103281B1 (en) * | 2019-08-09 | 2020-04-23 | 주식회사 딥노이드 | Ai based assistance diagnosis system for diagnosing cerebrovascular disease |
| CN111627553A (en) * | 2020-05-26 | 2020-09-04 | 四川大学华西医院 | Method for constructing individualized prediction model of first-onset schizophrenia |
| CN112353381A (en) * | 2020-11-24 | 2021-02-12 | 杭州冉曼智能科技有限公司 | Alzheimer's disease comprehensive diagnosis system based on multi-modal brain images |
| CN113012816A (en) * | 2021-04-12 | 2021-06-22 | 东北大学 | Brain partition risk prediction method and device, electronic equipment and storage medium |
| CN113538464A (en) * | 2021-07-22 | 2021-10-22 | 脑玺(苏州)智能科技有限公司 | Brain image segmentation model training method, segmentation method and device |
| CN113554663A (en) * | 2021-06-08 | 2021-10-26 | 浙江大学 | System for automatically analyzing dopamine transporter PET image based on CT structural image |
| CN113711271A (en) * | 2019-03-15 | 2021-11-26 | 豪夫迈·罗氏有限公司 | Deep convolutional neural network for tumor segmentation by positron emission tomography |
| CN113780421A (en) * | 2021-06-07 | 2021-12-10 | 广州天鹏计算机科技有限公司 | Brain PET image identification method based on artificial intelligence |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105212936A (en) * | 2015-09-09 | 2016-01-06 | 首都医科大学宣武医院 | Brain template generation method |
| KR102102255B1 (en) * | 2019-05-14 | 2020-04-20 | 주식회사 뷰노 | Method for aiding visualization of lesions in medical imagery and apparatus using the same |
-
2021
- 2021-12-25 CN CN202111605249.1A patent/CN114334130B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113711271A (en) * | 2019-03-15 | 2021-11-26 | 豪夫迈·罗氏有限公司 | Deep convolutional neural network for tumor segmentation by positron emission tomography |
| CN110335665A (en) * | 2019-04-01 | 2019-10-15 | 佛山原子医疗设备有限公司 | It is a kind of applied to medical image auxiliary diagnosis analysis to scheme to search drawing method and system |
| KR102103281B1 (en) * | 2019-08-09 | 2020-04-23 | 주식회사 딥노이드 | Ai based assistance diagnosis system for diagnosing cerebrovascular disease |
| CN111009324A (en) * | 2019-11-25 | 2020-04-14 | 东北大学 | Mild cognitive impairment auxiliary diagnosis system and method based on brain network multi-feature analysis |
| CN111627553A (en) * | 2020-05-26 | 2020-09-04 | 四川大学华西医院 | Method for constructing individualized prediction model of first-onset schizophrenia |
| CN112353381A (en) * | 2020-11-24 | 2021-02-12 | 杭州冉曼智能科技有限公司 | Alzheimer's disease comprehensive diagnosis system based on multi-modal brain images |
| CN113012816A (en) * | 2021-04-12 | 2021-06-22 | 东北大学 | Brain partition risk prediction method and device, electronic equipment and storage medium |
| CN113780421A (en) * | 2021-06-07 | 2021-12-10 | 广州天鹏计算机科技有限公司 | Brain PET image identification method based on artificial intelligence |
| CN113554663A (en) * | 2021-06-08 | 2021-10-26 | 浙江大学 | System for automatically analyzing dopamine transporter PET image based on CT structural image |
| CN113538464A (en) * | 2021-07-22 | 2021-10-22 | 脑玺(苏州)智能科技有限公司 | Brain image segmentation model training method, segmentation method and device |
Non-Patent Citations (1)
| Title |
|---|
| 周益琰 ; 丁广太 ; 宋安平 ; .一种在医学图像中挖掘非对称区域的方法.计算机辅助设计与图形学学报.2007,(第05期),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114334130A (en) | 2022-04-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114334130B (en) | Brain symmetry-based PET molecular image computer-aided diagnosis system | |
| US8774481B2 (en) | Atlas-assisted synthetic computed tomography using deformable image registration | |
| EP3226766B1 (en) | System and method for image calibration | |
| CN111008984B (en) | Automatic contour line drawing method for normal organ in medical image | |
| US8787648B2 (en) | CT surrogate by auto-segmentation of magnetic resonance images | |
| Chou et al. | 2D/3D image registration using regression learning | |
| CN108171738B (en) | Multi-modal medical image registration method based on brain function template | |
| US20160292864A1 (en) | Medical image data processing apparatus and method for determining the presence of an abnormality | |
| CN110246580B (en) | Cranial image analysis method and system based on neural network and random forest | |
| WO2016047683A1 (en) | Medical image display processing method, medical image display processing device, and program | |
| Malsch et al. | An enhanced block matching algorithm for fast elastic registration in adaptive radiotherapy | |
| Slomka et al. | Automated three-dimensional quantification of myocardial perfusion and brain SPECT | |
| JP4921472B2 (en) | Knowledge-based input region of interest definition for pharmacokinetic modeling | |
| CN113177991A (en) | Method for correcting scattering artifacts in CBCT (cone beam computed tomography) based on planned CT (computed tomography) | |
| Liu et al. | Robust cortical thickness morphometry of neonatal brain and systematic evaluation using multi-site MRI datasets | |
| JP2005230456A (en) | Method and apparatus for supporting diagnosis of cerebropathy | |
| JP6705528B2 (en) | Medical image display processing method, medical image display processing apparatus and program | |
| CN111161182A (en) | MR structure information constrained non-local mean guided PET image partial volume correction method | |
| CN114881914A (en) | System and method for determining three-dimensional functional liver segment based on medical image | |
| Dempere-Marco et al. | Analysis of visual search patterns with EMD metric in normalized anatomical space | |
| US20230076352A1 (en) | Systems and methods for image processing | |
| KR20210079133A (en) | Template-based Hippocampus Subfield Atrophy Analysis in Alzheimer's Disease and Normal Aging | |
| EP3624058A1 (en) | Method and system of analyzing symmetry from image data | |
| CN110858412A (en) | Image registration-based heart coronary artery CTA model establishing method | |
| Li et al. | 3D intersubject warping and registration of pulmonary CT images for a human lung model |
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 |