CN110533664B - Cranial nerve automatic segmentation method based on large sample data drive - Google Patents
Cranial nerve automatic segmentation method based on large sample data drive Download PDFInfo
- Publication number
- CN110533664B CN110533664B CN201910680807.7A CN201910680807A CN110533664B CN 110533664 B CN110533664 B CN 110533664B CN 201910680807 A CN201910680807 A CN 201910680807A CN 110533664 B CN110533664 B CN 110533664B
- Authority
- CN
- China
- Prior art keywords
- fiber
- cranial nerve
- registration
- sample
- clustering
- 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
- 210000003792 cranial nerve Anatomy 0.000 title claims abstract description 52
- 238000000034 method Methods 0.000 title claims abstract description 33
- 230000011218 segmentation Effects 0.000 title claims abstract description 19
- 239000000835 fiber Substances 0.000 claims abstract description 169
- 238000003384 imaging method Methods 0.000 claims abstract description 26
- 210000000133 brain stem Anatomy 0.000 claims abstract description 14
- 210000001638 cerebellum Anatomy 0.000 claims abstract description 14
- 210000004556 brain Anatomy 0.000 claims abstract description 13
- 238000013507 mapping Methods 0.000 claims description 16
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 6
- 210000005013 brain tissue Anatomy 0.000 claims description 6
- 230000009466 transformation Effects 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 4
- 239000013598 vector Substances 0.000 claims description 4
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 210000003901 trigeminal nerve Anatomy 0.000 claims description 3
- 238000002598 diffusion tensor imaging Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 3
- 210000005036 nerve Anatomy 0.000 description 3
- 210000004885 white matter Anatomy 0.000 description 3
- 238000009792 diffusion process Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 208000012661 Dyskinesia Diseases 0.000 description 1
- 208000037273 Pathologic Processes Diseases 0.000 description 1
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 210000003484 anatomy Anatomy 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001055 chewing effect Effects 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000002059 diagnostic imaging Methods 0.000 description 1
- 230000004424 eye movement Effects 0.000 description 1
- 210000003128 head Anatomy 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 238000002595 magnetic resonance imaging Methods 0.000 description 1
- 210000004400 mucous membrane Anatomy 0.000 description 1
- 210000003205 muscle Anatomy 0.000 description 1
- 210000003928 nasal cavity Anatomy 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000009054 pathological process Effects 0.000 description 1
- 230000011514 reflex Effects 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 230000035807 sensation Effects 0.000 description 1
- 230000008786 sensory perception of smell Effects 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000002560 therapeutic procedure Methods 0.000 description 1
- 210000001519 tissue Anatomy 0.000 description 1
- 230000016776 visual perception Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
-
- 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
- G06T2207/30016—Brain
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
A cranial nerve fiber automatic segmentation method based on large sample data drive is characterized in that a FreeSprofer tool is used for brain region segmentation, a deterministic tracking method based on spherical deconvolution (SD _ STREAM) is applied to carry out fiber tracking on a brainstem part and a cerebellum part on original DTI data, and on the basis of a cranial nerve fiber map obtained by a multi-sample registration and clustering method, sample data is mapped and registered to obtain a target cranial nerve fiber bundle. Compared with the common method for manually drawing the ROI and eliminating wrong fibers, the method eliminates subjectivity introduced by manual operation, ensures fiber bundle imaging certainty, reduces data errors, and can provide an efficient, accurate, stable and repeatable method for cranial nerve fiber segmentation.
Description
Technical Field
The invention relates to the field of medical imaging and neuroanatomy under computer graphics, in particular to a cranial nerve automatic segmentation method based on a fiber atlas.
Background
Cranial nerves are important functions of the human body 12 in controlling right and left paired nerves from the brain, such as the sense of smell, sight, and eye movement. 12 of the cranial nerves, except for the olfactory and sublingual nerves, the remaining 10 nerves are more likely to be damaged. Injury to cranial nerves can cause loss of general sensation in the skin of the head and face, as well as mucous membranes of the tongue, mouth, and nasal cavity, dyskinesia of affected chewing muscles, loss of corneal reflex, and the like. In cranial nerve therapy, fiber imaging techniques are used to make up for the clinical examination's deficiencies in the diagnosis of pathological processes. The method currently commonly used for cranial nerve fiber imaging is Diffusion Tensor Imaging (DTI). DTI is a special form of Magnetic Resonance Imaging (MRI) that maps the appearance of white matter fiber tracts in terms of the direction of water molecule movement. Meanwhile, high-angle resolution diffusion imaging (HARDI) technologies, such as Q sphere imaging (QBI), Constrained Sphere Deconvolution (CSD) and Diffusion Spectrum Imaging (DSI), are available to make up for the limitations of DTI on angular resolution.
In conventional cranial nerve fiber imaging, a practitioner with anatomical knowledge is required to draw a region of interest (ROI), i.e., the brain structure tissue in which cranial nerves are present, in each sample magnetic resonance image prior to fiber tracking. And taking the ROI as a seed point, performing fiber tracking on the sample by using a fiber tracking algorithm such as probabilistic tracking and deterministic tracking, wherein the tracked fiber result needs professionals to manually remove fiber bundles which do not belong to cranial nerves according to anatomical knowledge so as to obtain the individual cranial nerve fiber bundles. From the above, in the past cranial nerve fiber imaging process, because a large number of operations need to be completed manually, more uncertain factors are introduced, and the factors include: the judgment standard of the ROI, the familiarity of software and the like can greatly influence the result, and the instability of the imaging result is increased.
The fibrous atlas refers to an anatomy-based white matter atlas of the brain, which is derived from large sample data and can achieve consistent white matter segmentation among different populations. The fiber map has the characteristic of reusability, single data after the fiber map is established can be used for obtaining corresponding clustering labels, and repetitive operations of manually drawing ROI and manually removing fibers by medical workers can be well avoided. The automatic segmentation based on the large sample data driving is added in cranial nerve fiber imaging, so that the interference of human factors can be eliminated, and a large amount of complex manual operation is reduced. The method has important significance in improving the result accuracy of cranial nerve imaging and reducing resource investment.
Disclosure of Invention
The existing imaging technology depends on the operation of medical workers too much, has strong subjectivity, and can not ensure that the ROI sketched each time is accurate, thereby easily causing fiber bundle imaging uncertainty and data error. In order to overcome the defects, the invention provides a cranial nerve automatic segmentation method based on large sample data by combining a fiber atlas.
The technical scheme adopted by the invention for solving the technical problems is as follows:
a cranial nerve automatic segmentation method based on large sample data driving comprises the following steps:
step one, preprocessing data;
each DTI data is subjected to denoising, eddy current and motion correction, so that potential artifacts are avoided;
step two, brain tissue segmentation;
using a recon-all command in a FreeSenr tool to complete part or all of the process of rebuilding the FreeSenr cortex; before starting, the structure image is stored in a directory with hierarchy, the terminal is opened under the file, and the following are input: tcsh; inputting related parameters, designating a piece in the DICOM sequence as a subject name, designating a folder stored by the subject, obtaining a division result in the designated folder after the division is finished, and converting the division result into nii files;
step three, fiber tracking
Performing fiber tracking of a brain stem and a cerebellum part on original DTI data by applying a deterministic tracking method based on spherical deconvolution (SD _ STREAM), taking a fiber direction distribution (FOD) image represented on the basis of a spherical harmonic function (SH) as an input, sampling local (trilinear interpolation) FOD in each streamline step, performing Newton optimization on a sphere from the current streamline tangential direction so as to locate the direction of the nearest FOD amplitude peak value, and performing fiber tracking on the brain stem and the cerebellum in the whole brain range by taking the brain stem and the cerebellum as seed points;
step four, fiber registration mapping
The creation of the atlas and the measurement of the fiber bundle statistics both require mapping a plurality of datasets to a common coordinate system through registration, performing initial alignment or standardization of fiber bundle imaging, applying an unbiased multi-sample affine and B-spline-based fiber bundle imaging registration method, and a B-spline model is used to realize non-rigid group registration to improve the spatial correspondence quality.
Further, in the fourth step, the fiber registration mapping comprises the following steps:
4.1, multi-sample registration;
performing multi-sample registration when generating a fiber atlas, performing entropy-based registration in a multi-scale mode based on paired fiber track distance, performing affine transformation at first, performing non-rigid b-spline transformation, and aligning fibers from a plurality of samples; converting the input variable length fiber traces into a fixed length representation, each fiber being represented by an end point, a midpoint, and two intermediate points; randomly sampling 50000 fibers from the fiber bundle imaging of each sample for fiber bundle registration, wherein mapping parameters are constrained, an initial step length is given, and convergence is determined when a final step length is lower than a set threshold value;
4.2, clustering trigeminal nerve fiber bundle atlas;
in the feature space, all fibers compute the feature vector through the similarity matrix and are represented as points in the feature space. Points near the fiber generally correspond to similar fiber trajectories. After group fiber registration, each fiber is converted to a point in feature space, clustering is performed to automatically generate a fiber map. The feature space represents the fiber bundle in terms of similarity between fibers. Each fiber from each sample image is assigned to the closest fiber map cluster in feature space, and if the fiber similarity exceeds 2 standard deviations from the average fiber similarity, the fiber is determined to be an abnormal fiber and removed from the map;
the atlas clustering process is divided into two classifications, 1) positioning cranial nerve fiber bundles, and performing first classification and positioning on fiber data to cranial nerve fiber bundles in order to quickly and accurately mark cranial nerves, wherein the first clustering result confirms the classification of cranial nerve fibers in the clustering result through brain anatomical structure comparison, and the types of the cranial nerve fibers are obtained by tracking based on brainstem and cerebellum as seed points; 2) performing secondary classification, wherein the fasciculus containing cranial nerve fibers also contains other fasciculus except the fasciculus of cranial nerve fibers, and the fasciculus of cranial nerve fibers can be obtained by performing single secondary clustering on the fasciculus of the class;
4.3, new sample registration mapping;
and after the fiber atlas is established, the fiber atlas is used for new sample automatic segmentation fiber bundle imaging, each new fiber sample is represented in the feature space for initial clustering, and then a clustering label is distributed according to the nearest clustering centroid. Firstly, segmenting brain tissues to obtain brainstem and cerebellum parts of each group of data as seed points to perform deterministic tracking in the whole brain range, secondly, performing registration mapping on each group of fiber data and the map of the first classification, and finally, extracting corresponding fiber bundles to map the map of the second classification to obtain cranial nerve fiber bundles of a new sample.
The invention has the beneficial effects that: effectively realizes automatic cranial nerve segmentation.
Drawings
FIG. 1 is a flow chart of the implementation steps of the method.
Fig. 2 is a process of creating a cranial nerve fiber bundle map.
Detailed Description
The present invention is further described below.
Referring to fig. 1 and 2, a cranial nerve fiber automatic segmentation method based on large sample data driving comprises the following steps:
step one, preprocessing data;
each DTI data is subjected to denoising, eddy current and motion correction, so that potential artifacts are avoided;
step two, brain tissue segmentation;
using a recon-all command in a FreeSenr tool to complete part or all of the process of rebuilding FreeSenr cortex, storing a structural image into a hierarchical directory before starting, opening a terminal under a file, and inputting: tcsh; inputting related parameters, designating a piece in the DICOM sequence as a subject name, designating a folder stored by the subject, obtaining a division result in the designated folder after the division is finished, and converting the division result into nii files;
step three, fiber tracking
The method is characterized in that a deterministic tracking method based on spherical deconvolution (SD _ STREAM) is applied to perform fiber tracking on the brain stem and cerebellum parts of original DTI data, a fiber direction distribution (FOD) image represented on the basis of a spherical harmonic function (SH) is used as an input, local (trilinear interpolation) FOD is sampled in each streamline step, a sphere is subjected to Newton optimization from the tangential direction of the current streamline so as to locate the direction of the nearest FOD amplitude peak, and the brain stem and the cerebellum are used as seed points to perform fiber tracking in the whole brain range.
Step four, fiber registration mapping
Both the creation of the atlas and the measurement of the fiber bundle statistics require initial alignment or normalization of the fiber bundle imaging by mapping multiple datasets to a common coordinate system by registration. The invention applies an unbiased multi-sample affine and B-spline-based fiber bundle imaging registration method, and a B-spline model is used for realizing non-rigid grouping registration to improve the space corresponding quality.
In the fourth step, the fiber registration mapping comprises the following steps:
4.1, multi-sample registration;
performing multi-sample registration when generating a fiber atlas, performing entropy-based registration in a multi-scale mode based on paired fiber track distance, performing affine transformation at first, performing non-rigid b-spline transformation, and aligning fibers from a plurality of samples; converting the input variable length fiber traces into a fixed length representation, each fiber being represented by an end point, a midpoint, and two intermediate points; randomly sampling 50000 fibers from the fiber bundle imaging of each sample for fiber bundle registration, wherein mapping parameters are constrained, an initial step length is given, and convergence is determined when a final step length is lower than a set threshold value;
4.2, clustering trigeminal nerve fiber bundle atlas;
in the feature space, all fibers compute the feature vector through the similarity matrix and are represented as points in the feature space. Points near the fiber generally correspond to similar fiber trajectories. After group fiber registration, each fiber is converted to a point in feature space, clustering is performed to automatically generate a fiber map. The feature space represents the fiber bundle in terms of similarity between fibers. Each fiber from each sample image is assigned to the closest fiber map cluster in feature space, and if the fiber similarity exceeds 2 standard deviations from the average fiber similarity, the fiber is determined to be an abnormal fiber and removed from the map;
the atlas clustering process is divided into two classifications, 1) positioning cranial nerve fiber bundles, and performing first classification and positioning on fiber data to cranial nerve fiber bundles in order to quickly and accurately mark cranial nerves, wherein the first clustering result confirms the classification of cranial nerve fibers in the clustering result through brain anatomical structure comparison, and the types of the cranial nerve fibers are obtained by tracking based on brainstem and cerebellum as seed points; 2) performing secondary classification, wherein the fasciculus containing cranial nerve fibers also contains other fasciculus except the fasciculus of cranial nerve fibers, and the fasciculus of cranial nerve fibers can be obtained by performing single secondary clustering on the fasciculus of the class;
4.3, new sample registration mapping;
and after the fiber atlas is established, the fiber atlas is used for new sample automatic segmentation fiber bundle imaging, each new fiber sample is represented in the feature space for initial clustering, and then a clustering label is distributed according to the nearest clustering centroid. Firstly, segmenting brain tissues to obtain brainstem and cerebellum parts of each group of data as seed points to perform deterministic tracking in the whole brain range, secondly, performing registration mapping on each group of fiber data and the map of the first classification, and finally, extracting corresponding fiber bundles to map the map of the second classification to obtain cranial nerve fiber bundles of a new sample.
Claims (1)
1. A cranial nerve fiber automatic segmentation method based on large sample data driving is characterized in that: the method comprises the following steps:
step one, preprocessing data;
denoising, eddy current and motion correction are carried out on each DTI data;
step two, brain tissue segmentation;
using a recon-all command in a FreeSenr tool to complete part or all of the process of rebuilding FreeSenr cortex, storing a structural image into a hierarchical directory before starting, opening a terminal under a file, and inputting: tcsh; inputting related parameters, designating a piece in the DICOM sequence as a subject name, designating a folder stored by the subject, obtaining a division result in the designated folder after the division is finished, and converting the division result into nii files;
step three, fiber tracking
Performing fiber tracking on a brainstem and cerebellum part on original DTI data by using a deterministic tracking method based on spherical deconvolution, wherein a fiber direction distribution FOD image represented on the basis of a spherical harmonic function SH is used as input in the method, local FOD is sampled in each streamline step, and Newton optimization is performed on a spheroid from the tangential direction of the current streamline so as to position the direction of the nearest FOD amplitude peak value, and the brainstem and the cerebellum are used as seed points to perform fiber tracking in the whole brain range;
step four, fiber registration mapping
The establishment of the atlas and the measurement of the fiber bundle statistics both need to map a plurality of data sets to a common coordinate system through registration, carry out the initial alignment or standardization of the fiber bundle imaging, apply an unbiased multi-sample affine and B-spline-based fiber bundle imaging registration method, and the B-spline model is used for realizing the non-rigid grouping registration to improve the corresponding quality of the space;
in the fourth step, the fiber registration mapping comprises the following steps:
4.1, multi-sample registration;
performing multi-sample registration when generating a fiber atlas, performing entropy-based registration in a multi-scale mode based on paired fiber track distance, performing affine transformation at first, performing non-rigid b-spline transformation, and aligning fibers from a plurality of samples; converting the input variable length fiber traces into a fixed length representation, each fiber being represented by an end point, a midpoint, and two intermediate points; randomly sampling 50000 fibers from the fiber bundle imaging of each sample for fiber bundle registration, wherein mapping parameters are constrained, an initial step length is given, and convergence is determined when a final step length is lower than a set threshold value;
4.2, clustering trigeminal nerve fiber bundle atlas;
in the feature space, all fibers calculate feature vectors through a similarity matrix and represent the feature vectors as points in the feature space, the points near the fibers correspond to fiber tracks, after group fiber registration, each fiber is converted into the points in the feature space, clustering is carried out to automatically generate a fiber map, the feature space represents fiber bundles according to the similarity between the fibers, each fiber from each sample imaging is distributed to the nearest fiber map cluster in the feature space, if the fiber similarity exceeds 2 standard deviations relative to the average fiber similarity, the fiber is judged to be abnormal fiber and removed from the map;
the atlas clustering process is divided into two classifications, 1) positioning cranial nerve fiber bundles, and performing first classification and positioning on fiber data to cranial nerve fiber bundles in order to quickly and accurately mark cranial nerves, wherein the first clustering result confirms the classification of cranial nerve fibers in the clustering result through brain anatomical structure comparison, and the types of the cranial nerve fibers are obtained by tracking based on brainstem and cerebellum as seed points; 2) performing secondary classification, wherein the fasciculus containing cranial nerve fibers also contains other fasciculus except the fasciculus of cranial nerve fibers, and the fasciculus of cranial nerve fibers can be obtained by performing single secondary clustering on the fasciculus of the class;
4.3, new sample registration mapping;
after the fiber atlas is established, the new fiber atlas is used for automatic segmentation fiber bundle imaging of new samples, each new fiber sample is expressed in a feature space which is initially clustered, then a clustering label is distributed according to the nearest clustering center of mass, firstly, brain tissues are segmented to obtain brainstem and cerebellum parts of each group of data as seed points to perform deterministic tracking in the whole brain range, secondly, each group of fiber data and the atlas of the first classification are registered and mapped, and finally, corresponding fiber bundles are extracted and mapped to the atlas of the second classification to obtain cranial nerve fiber bundles of the new samples.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910680807.7A CN110533664B (en) | 2019-07-26 | 2019-07-26 | Cranial nerve automatic segmentation method based on large sample data drive |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910680807.7A CN110533664B (en) | 2019-07-26 | 2019-07-26 | Cranial nerve automatic segmentation method based on large sample data drive |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110533664A CN110533664A (en) | 2019-12-03 |
| CN110533664B true CN110533664B (en) | 2022-05-03 |
Family
ID=68661873
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910680807.7A Active CN110533664B (en) | 2019-07-26 | 2019-07-26 | Cranial nerve automatic segmentation method based on large sample data drive |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110533664B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111161226A (en) * | 2019-12-20 | 2020-05-15 | 西北工业大学 | Method for uniformly segmenting cerebral cortex surface based on spectral clustering algorithm |
| CN111710010A (en) * | 2020-06-04 | 2020-09-25 | 浙江工业大学 | Cranial nerve automatic imaging method based on deep network learning |
| CN111739580B (en) * | 2020-06-15 | 2021-11-02 | 西安电子科技大学 | Brain white matter fiber bundle clustering method based on fiber midpoint and end points |
| CN111986101B (en) * | 2020-07-09 | 2024-06-18 | 浙江工业大学 | Cerebrovascular map construction method |
| CN111951279B (en) * | 2020-08-05 | 2024-04-23 | 中国科学院深圳先进技术研究院 | Brain white matter region segmentation method and device of brain ultrasonic image and electronic equipment |
| CN113591616B (en) * | 2021-07-14 | 2024-02-13 | 华中科技大学 | Nerve fiber reconstruction method and system based on foreground point clustering |
| CN116109571B (en) * | 2022-12-22 | 2024-04-26 | 浙江大学 | Automatic fiber bundle reconstruction method and system for facial auditory nerve |
| CN116542997B (en) * | 2023-07-04 | 2023-11-17 | 首都医科大学附属北京朝阳医院 | Magnetic resonance image processing method and device and computer equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103093455A (en) * | 2012-12-21 | 2013-05-08 | 西北工业大学 | Diffusion tensor imaging white matter fiber clustering method |
| CN104523275A (en) * | 2014-12-25 | 2015-04-22 | 西安电子科技大学 | Construction method for health people white matter fiber tract atlas |
| CN106204562A (en) * | 2016-07-04 | 2016-12-07 | 西安交通大学 | A kind of method of the arched roof white matter segmentation merged based on fMRI Yu DTI |
| WO2018237011A1 (en) * | 2017-06-20 | 2018-12-27 | University Of Louisville Research Foundation, Inc. | Segmentation of retinal blood vessels in optical coherence tomography angiography images |
| CN109978871A (en) * | 2019-03-30 | 2019-07-05 | 西安电子科技大学 | Merge the fibre bundle screening technique of probabilistic type and the tracking of deterministic type fibre bundle |
-
2019
- 2019-07-26 CN CN201910680807.7A patent/CN110533664B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103093455A (en) * | 2012-12-21 | 2013-05-08 | 西北工业大学 | Diffusion tensor imaging white matter fiber clustering method |
| CN104523275A (en) * | 2014-12-25 | 2015-04-22 | 西安电子科技大学 | Construction method for health people white matter fiber tract atlas |
| CN106204562A (en) * | 2016-07-04 | 2016-12-07 | 西安交通大学 | A kind of method of the arched roof white matter segmentation merged based on fMRI Yu DTI |
| WO2018237011A1 (en) * | 2017-06-20 | 2018-12-27 | University Of Louisville Research Foundation, Inc. | Segmentation of retinal blood vessels in optical coherence tomography angiography images |
| CN109978871A (en) * | 2019-03-30 | 2019-07-05 | 西安电子科技大学 | Merge the fibre bundle screening technique of probabilistic type and the tracking of deterministic type fibre bundle |
Non-Patent Citations (2)
| Title |
|---|
| "Automatic Tractography Segmentation Using a High-Dimensional White Matter Atlas";Lauren J. O"Donnell等;《IEEE Transactions on Medical Imaging》;20071130;第26卷(第11期);第1562-1575页 * |
| "BOLD和DTI图像在神经导航中的配准、纤维束示踪及临床应用";何黎民等;《中国医学影像技术》;20070430;第23卷(第4期);第495-499页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110533664A (en) | 2019-12-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110533664B (en) | Cranial nerve automatic segmentation method based on large sample data drive | |
| CN111931811B (en) | Calculation method based on super-pixel image similarity | |
| CN111739000B (en) | System and device for improving left ventricle segmentation accuracy of multiple cardiac views | |
| CN105205810B (en) | MR image hippocampus dividing method based on distance field fusion | |
| CN111161241B (en) | Liver image identification method, electronic equipment and storage medium | |
| CN107680107B (en) | Automatic segmentation method of diffusion tensor magnetic resonance image based on multiple maps | |
| Liu et al. | Automatic whole heart segmentation using a two-stage u-net framework and an adaptive threshold window | |
| CN110097557B (en) | Medical image automatic segmentation method and system based on 3D-UNet | |
| Kim et al. | Multivariate hippocampal subfield analysis of local MRI intensity and volume: application to temporal lobe epilepsy | |
| JP7092431B2 (en) | Methods, image processing devices and storage media for determining the median sagittal plane with magnetic resonance images | |
| CN110472737A (en) | Training method, device and the magic magiscan of neural network model | |
| CN110993065B (en) | Brain tumor keyhole surgery path planning method based on image guidance | |
| CN101504727B (en) | Image processing apparatus and method thereof | |
| CN112184720B (en) | Method and system for segmenting internal rectus muscle and optic nerve of CT image | |
| Rejusha et al. | Artificial MRI image generation using deep convolutional GAN and its comparison with other augmentation methods | |
| Skibbe et al. | PAT—Probabilistic axon tracking for densely labeled neurons in large 3-D micrographs | |
| CN110570430B (en) | Orbital bone tissue segmentation method based on volume registration | |
| Smith et al. | Automated torso contour extraction from clinical cardiac MR slices for 3D torso reconstruction | |
| CN116310114A (en) | Coronary artery point cloud data processing optimization method of multitask label model | |
| CN110415246A (en) | A kind of analysis method of stomach fat ingredient | |
| Ribbens et al. | SPARC: unified framework for automatic segmentation, probabilistic atlas construction, registration and clustering of brain MR images | |
| CN114782371A (en) | Brain area automatic segmentation method based on artificial intelligence | |
| CN112233058A (en) | Method for detecting lymph nodes in head and neck CT image | |
| Melonakos et al. | Locally-constrained region-based methods for dw-mri segmentation | |
| CN117934689B (en) | Multi-tissue segmentation and three-dimensional rendering method for fracture CT image |
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 | ||
| CB03 | Change of inventor or designer information |
Inventor after: Zeng Qingrun Inventor after: Feng Yuanjing Inventor after: Chen Yukai Inventor after: Jin Er Inventor after: Tan Zhihao Inventor after: Li Siqi Inventor after: Qiu Xinyu Inventor before: Feng Yuanjing Inventor before: Chen Yukai Inventor before: Jin Er Inventor before: Tan Zhihao Inventor before: Zeng Qingrun Inventor before: Li Siqi Inventor before: Qiu Xinyu |
|
| CB03 | Change of inventor or designer information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |