CN108305279A - A kind of brain magnetic resonance image super voxel generation method of iteration space fuzzy clustering - Google Patents
A kind of brain magnetic resonance image super voxel generation method of iteration space fuzzy clustering Download PDFInfo
- Publication number
- CN108305279A CN108305279A CN201711445476.6A CN201711445476A CN108305279A CN 108305279 A CN108305279 A CN 108305279A CN 201711445476 A CN201711445476 A CN 201711445476A CN 108305279 A CN108305279 A CN 108305279A
- Authority
- CN
- China
- Prior art keywords
- magnetic resonance
- voxel
- resonance image
- seed point
- space
- 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.)
- Granted
Links
Classifications
-
- 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
- G06T7/344—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods involving models
-
- 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
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30016—Brain
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (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
The invention discloses a kind of super voxel generation methods of the brain magnetic resonance image of iteration space fuzzy clustering, include the following steps：Firstly, since human brain topological structure having the same, one group of seed pattern is obtained from the big Typical AVM template based on group；Secondly, for the influence of exclusive segment volume effect, it is proposed that voxel is distributed to each seed and generates super voxel by a kind of iteration space fuzzy clustering algorithm.The present invention can preferably be applied to brain magnetic resonance image, generate the super voxel of effective brain magnetic resonance image.
Description
Technical field
The present invention relates to a kind of super voxel generation methods of the brain magnetic resonance image of iteration space fuzzy clustering, belong to number
Image domains.
Background technology
Super voxel technology is that the voxel with high redundancy feature is gathered into the process of significant homogeneous area.And biography
The image procossing basic unit voxel of system is compared, and the image analysis based on a certain number of super voxels and processing can obtain preferably
Effect, while efficiency can also be greatly improved.The regional area of brain magnetic resonance image is smooth, this to be suitble to total to brain magnetic
The image that shakes carries out super voxel segmentation.Recently, super voxel technology is increasingly being applied to brain magnetic resonance image analysis,
And fairly good performance is shown in some applications, such as tumor-localizing and segmentation, tissue segmentation, image registration and function point
Group etc..Therefore, the super voxel dividing method of good brain magnetic resonance MRI image to the analysis of follow-up brain MRI image to closing weight
It wants.
Brain MRI image has unique property.First, human brain has complicated internal structure, including several sizes are not
Minor structure same, complex shape degree is different.It organizes generally be divided into cerebrospinal fluid, white matter and white matter in brain, these three groups
It knits with complicated boundary shape and topological structure.Seed point is generated using traditional uniform sampling that carried out in lattice structure
Method may generate the super voxel of non-uniform brain magnetic resonance image.Secondly as lack of resolution in magnetic resonance image
It limits, the loss of contrast between adjacent tissue causes include Various Tissues, i.e. partial volume effect in a voxel.It passes
The hard clustering method for the one or the other that the super voxel generating algorithm for natural image of system uses clusters voxel.Make
Super voxel segmentation is carried out to brain magnetic resonance image with the algorithm clustered firmly, caused result, which may be in a super voxel, includes
Various Tissues (including grey matter, white matter and cerebrospinal fluid) so that the boundary fitting ability of super voxel is excessively poor.
Currently, have existed the super voxel method developed for natural image, for example, simple linear alternative manner, based on figure
The super voxel generating algorithm such as method, mean shift, and be applied to the numerous areas of computer vision, achieve fairly good
Result.However, in brain magnetic resonance image analysis, it is still to have to generate suitable super voxel for brain magnetic resonance image
Challenge.It is merely clearly inappropriate with the existing algorithm generation super voxel of brain MRI image, because these are directed to certainly
Right image algorithm has ignored the special nature of brain magnetic resonance image.
Invention content
Existing super voxel generating algorithm is to be directed to natural image mostly, cannot access ideal brain magnetic resonance figure
As super voxel, the present invention provides a kind of the big of iteration space fuzzy clustering to generate the super voxel of ideal brain magnetic resonance image
The super voxel generation method of brain magnetic resonance image.
The present invention uses following technical scheme to solve above-mentioned technical problem：
The present invention provides a kind of super voxel generation method of brain magnetic resonance image of iteration space fuzzy clustering, including following
Step：
Step 1, seed pattern is built from the brain magnetic resonance template image based on group, is subsequently projected to individual space life
At the seed point of individual space, specifically include：
Step 1, seed pattern is built from the brain magnetic resonance template image based on group, is subsequently projected to individual space life
At the seed point of individual space, specifically include：
1-1, in the N number of seed point of brain area domain uniform sampling of brain magnetic resonance template image；
1-2 calculates the distance between the seed point of each voxel and step 1-1 on brain magnetic resonance template image, wherein
The distance between a-th of voxel and b-th of seed point D (a, b)=dI(a,b)+λdS(a, b), dI(a, b) be a-th voxel with
Gray homogeneity between b-th of seed point, dSThe space length of (a, b) between a-th of voxel and b-th of seed point, λ are sky
Between distance and Gray homogeneity weight；
Two seed points with minimum range are merged into a new seed by 1-3 every time using Hierarchical clustering methods
Point generates seeding template；
1-4, in such a way that brain magnetic resonance template image is registrated to individual magnetic resonance image, by corresponding seed point
Template projects to the space of individual magnetic resonance image, generates the seed point in individual magnetic resonance image space；
Step 2, using a kind of method of iteration space fuzzy clustering, the voxel for calculating individual magnetic resonance image is nearest with it
Fuzzy membership between K seed generates super voxel, specifically includes：
2-1 initializes the seed point in the individual magnetic resonance image space in step 1-4 with vector, wherein individual
The initialization vector of j-th of seed point in magnetic resonance image space(xj,yj,zj) it is individual magnetic resonance
The coordinate of j-th of seed point of image space,It is j-th of seed point and its 3 × 3 × 3 neighbours in individual magnetic resonance image space
The average value of the voxel intensities of all voxels in domain；
Each voxel in individual magnetic resonance image K seed point nearest with it is carried out Fuzzy Correlation, and counted by 2-2
Calculate the fuzzy membership between each voxel and its K nearest seed：
Wherein,Indicate i-th of voxel in individual magnetic resonance image and seed point skBetween fuzzy membership, D (q,
sk) indicate i-th of voxel in individual magnetic resonance image and seed point skThe distance between, m indicates that the fuzzy of fuzzy membership adds
Weigh index, sk,st∈ S, S={ s1,s2,…,sK, S indicates nearest K kinds of i-th of voxel in individual magnetic resonance image
The set of son point, k, t=1,2 ..., K；
2-3, the fuzzy membership calculated using step 2-2 are updated the seed point in individual magnetic resonance image space,
It obtains：
Wherein, NjIndicate the quantity with the voxel of j-th of seed point Fuzzy Correlation in individual magnetic resonance image space, ujrTable
Show j-th of seed point and the fuzzy membership with r-th of voxel of its Fuzzy Correlation, v in individual magnetic resonance image spacer=
[xr,yr,zr,Ir]T, (xr,yr,zr) indicate r-th of body with j-th of seed point Fuzzy Correlation in individual magnetic resonance image space
The coordinate of element, IrIndicate the gray value with r-th of voxel of j-th of seed point Fuzzy Correlation in individual magnetic resonance image space；
2-4 carries out being iterated the seed point in individual magnetic resonance image space update according to step 2-2 and 2-3, and fixed
Adopted residual error stops iteration when residual error is less than given threshold, and each super voxel is distributed to maximum fuzzy person in servitude
The seed point of category degree is to generate the super voxel of individual brain magnetic resonance image.
Step 3, after determining super voxel number, the super voxel of brain magnetic resonance image is generated by step 2.
As the further technical solution of the present invention, the Gray homogeneity d between a-th of voxel and b-th of seed pointI(a,
B)=| Ia-Ib|, wherein IbAnd IaThe image pixel intensities of b-th of seed point and a-th of voxel are indicated respectively.
As the further technical solution of the present invention, the space length between a-th of voxel and b-th of seed pointWherein (xb,yb,zb) and (xa,ya,za) b-th kind is indicated respectively
The coordinate of son point and a-th of voxel.
As the further technical solution of the present invention, i-th of voxel in individual magnetic resonance image and seed point skBetween
Distance D (i, sk)=dI(i,sk)+λdS(i,sk), dI(i,sk) indicate i-th of voxel and seed in individual magnetic resonance image
Point skBetween Gray homogeneity, dS(i,sk) indicate i-th of voxel in individual magnetic resonance image and seed point skBetween space
Distance.
As the further technical solution of the present invention, the individual magnetic resonance figure after being updated defined in step 2-4 and before update
Euclidean distance between the seed point of image space is residual error.
As the further technical solution of the present invention, N=10000 in step 1-1.
The present invention has the following technical effects using above technical scheme is compared with the prior art：The invention discloses one
The super voxel generation method of brain magnetic resonance image of kind iteration space fuzzy clustering, firstly, since human brain is having the same
Topological structure obtains one group of seed pattern from the big Typical AVM template based on group；Secondly, for exclusive segment volume effect
It influences, it is proposed that voxel is distributed to each seed and generates super voxel by a kind of iteration space fuzzy clustering algorithm.The present invention can be compared with
It is applied to brain magnetic resonance image well, generates the super voxel of effective brain magnetic resonance image.
Description of the drawings
Fig. 1 is flow chart of the method for the present invention.
Fig. 2 is the implementation process diagram of the present invention.
Fig. 3 is brain magnetic resonance image.
Fig. 4 be proposition method of the present invention super voxel generate result, wherein (a) for super voxel number be 500 as a result,
(b) be super voxel number it is 1000 as a result, (c) being result that super voxel number is 2000.
Specific implementation mode
Technical scheme of the present invention is described in further detail below in conjunction with the accompanying drawings：
The present invention provides a kind of super voxel generation method of brain magnetic resonance image of iteration space fuzzy clustering, such as Fig. 1 and 2
It is shown, first, one group of seed pattern is built from the brain magnetic resonance image template based on group；Then a kind of iteration space is used
The method of fuzzy clustering calculates the fuzzy membership between voxel and its nearest K seed, generates super voxel；Finally, in determination
After super voxel number, super voxel is generated according to step 1 and step 2.
First, the generation of seed pattern is specially：Seed pattern first is built from the big Typical AVM template based on group, then
Individual space is projected to generate reliable seed, including following 4 steps in the generation of super voxel：
(1-1) equably samples N number of seed point in the brain area domain of brain magnetic resonance template image.
(1-2) calculates each space length between voxel and the seed point of step 1-1 on brain magnetic resonance template image
dSWith Gray homogeneity dI, wherein dI、dSCalculation formula it is as follows：
dI(a, b)=| Ia-Ib| (1)
In above formula, dIThe Gray homogeneity of (a, b) between a-th of voxel and b-th of seed point, IbAnd IaB is indicated respectively
The image pixel intensities of a seed point and a-th of voxel.
In above formula, dSThe space length of (a, b) between a-th of voxel and b-th of seed point, (xb,yb,zb) and (xa,
ya,za) coordinate of b-th of seed point and a-th of voxel is indicated respectively.
By space length dSWith Gray homogeneity dICalculate distance D：
D (a, b)=dI(a,b)+λdS(a,b) (3)
In above formula, D (a, b) is the distance between a-th of voxel and b-th of seed point, λ representation space distances dSWith gray scale
Distance dIWeight.
The distance D that (1-3) is calculated based on step 1-2, using existing disclosed Hierarchical clustering methods, every time to having minimum
Two seeds of distance are merged into a new seed, generate seeding template.Phase will be used during merging seed
The space length of adjacent seed and the weighted array of voxel intensities are measured as the distance between seed.
(1-4) in such a way that brain magnetic resonance template image is registrated to individual magnetic resonance image, by corresponding seed
Template projects to the space of individual magnetic resonance image, generates the seed point in individual magnetic resonance image space.
Then, iteration space fuzzy clustering concrete operations are：
(2-1) by each seed point in step 1-4 with vector initialising, j-th kind of individual magnetic resonance image space
The initialization vector C of son pointjIt is defined as follows：
Wherein, (xj,yj,zj) be individual magnetic resonance image space j-th of seed point coordinate；It is individual magnetic resonance
The average value of the voxel intensities of all voxels, can inhibit in this way in j-th of seed point of image space and its 3 × 3 × 3 neighborhoods
The influence of noise improves the robustness of algorithm.
Each voxel in individual magnetic resonance image K seed nearest with it is carried out Fuzzy Correlation by (2-2).It calculates every
Fuzzy membership between a voxel and its K nearest seed, fuzzy membership calculation formula are as follows：
Wherein,Indicate i-th of voxel in individual magnetic resonance image and seed point skBetween fuzzy membership, m tables
Show the FUZZY WEIGHTED index of fuzzy membership, sk,st∈ S, S={ s1,s2,…,sK, S indicates i-th in individual magnetic resonance image
The set of the K nearest seed point of a voxel, k, t=1,2 ..., K.
D(i,sk) indicate i-th of voxel in individual magnetic resonance image and seed point skThe distance between, definition and step
1-2 is identical：
D(i,sk)=dI(i,sk)+λdS(i,sk) (6)
Wherein, dI(i,sk) indicate i-th of voxel in individual magnetic resonance image and seed point skBetween Gray homogeneity,
dS(i,sk) indicate i-th of voxel in individual magnetic resonance image and seed point skBetween space length.
(2-3) is updated seed point using the step 2-2 fuzzy memberships calculated, and more new formula is as follows：
Wherein, NjIt is the quantity indicated with the voxel of j-th of seed point Fuzzy Correlation in individual magnetic resonance image space, ujr
Indicate j-th of seed point and the fuzzy membership with r-th of voxel of its Fuzzy Correlation, v in individual magnetic resonance image spacerIt is
With the vector description of r-th of voxel of j-th of seed point Fuzzy Correlation in individual magnetic resonance image space (be similar to seed to
Amount description), it is represented by the following formula：
vr=[xr,yr,zr,Ir]T (8)
Wherein, (xr,yr,zr) indicate r-th of body with j-th of seed point Fuzzy Correlation in individual magnetic resonance image space
The coordinate of element, IrIndicate the gray value with r-th of voxel of j-th of seed point Fuzzy Correlation in individual magnetic resonance image space.
(2-4) carries out being iterated the seed point in individual magnetic resonance image space update according to step 2-2 and 2-3, and
Residual error is defined, stops iteration when residual error is less than given threshold, each super voxel is distributed to maximum fuzzy
The seed point of degree of membership is to generate the super voxel of individual brain magnetic resonance image.
Finally, after determining super voxel number, the seed point of individual space is generated by step 1, is generated by step 2
The super voxel of brain magnetic resonance image.
In the present invention, a large amount of seed is equably sampled in the brain area domain of brain image, uses the iteration space of the present invention
Fuzzy clustering algorithm generates seed.Then, these are obtained by a kind of hierarchical clustering algorithm keeping the adjoining property of space of cluster
Seed further merge generate seed pattern.During merging seed, the space length and voxel of neighboring seeds are used
The weighted array of intensity is measured as the distance between seed, and two seeds with minimum range are merged into a new kind
Son.In such a way that brain template magnetic resonance image is registrated to individual images, corresponding seed pattern is projected into individual sky
Between.Template image is registrated with single image in pairs to execute projection by using efficient Elastix tools.
Embodiment：
Below with BrainWeb18 data set data instances, to illustrate the brain magnetic of iteration space fuzzy clustering of the invention
The super voxel generating algorithm of resonance.
Experiment condition：It now chooses a computer to be tested, which is configured with Intel processors
(3.4GHz) and 10GB random access memory, 64 bit manipulation systems, programming language is Matlab (R2014a version).
Experimental data is the brain magnetic resonance image of BrainWeb18 data sets.Each MRI image includes 181 × 217 ×
The voxel that 181 sizes are 1mm × 1mm × 1mm.Experiment parameter is set as：K=6, λ=0.8, super voxel number are respectively set
For 500,1000 and 2000.Fig. 3 is BrainWeb18 brain MRI image artworks, and it is 500 that (a) in Fig. 4, which is super voxel number,
As a result, it is 1000 as a result, (c) in Fig. 4 is the result that super voxel number is 2000 that (b) in Fig. 4, which is super voxel number,.
The above, the only specific implementation mode in the present invention, but scope of protection of the present invention is not limited thereto, appoints
What is familiar with the people of the technology within the technical scope disclosed by the invention, it will be appreciated that expects transforms or replaces, and should all cover
Within the scope of the present invention, therefore, the scope of protection of the invention shall be subject to the scope of protection specified in the patent claim.
Claims (6)
1. a kind of super voxel generation method of the brain magnetic resonance image of iteration space fuzzy clustering, which is characterized in that including following
Step：
Step 1, seed pattern is built from the brain magnetic resonance template image based on group, is subsequently projected to individual space generation
The seed point in body space, specifically includes：
1-1, in the N number of seed point of brain area domain uniform sampling of brain magnetic resonance template image；
1-2 calculates the distance between the seed point of each voxel and step 1-1 on brain magnetic resonance template image, wherein a
The distance between a voxel and b-th of seed point D (a, b)=dI(a,b)+λdS(a, b), dI(a, b) is a-th of voxel and b
Gray homogeneity between a seed point, dSThe space length of (a, b) between a-th of voxel and b-th of seed point, λ are space
The weight of distance and Gray homogeneity；
Two seed points with minimum range are merged into a new seed point by 1-3 every time using Hierarchical clustering methods,
Generate seeding template；
1-4, in such a way that brain magnetic resonance template image is registrated to individual magnetic resonance image, by corresponding seeding template
The space of individual magnetic resonance image is projected to, the seed point in individual magnetic resonance image space is generated；
Step 2, using a kind of method of iteration space fuzzy clustering, voxel and its nearest K for calculating individual magnetic resonance image are a
Fuzzy membership between seed generates super voxel, specifically includes：
2-1 initializes the seed point in the individual magnetic resonance image space in step 1-4 with vector, wherein individual magnetic is total
Shake image space j-th of seed point initialization vector(xj,yj,zj) it is individual magnetic resonance image
The coordinate of j-th of seed point in space,It is in j-th of seed point and its 3 × 3 × 3 neighborhoods in individual magnetic resonance image space
The average value of the voxel intensities of all voxels；
Each voxel in individual magnetic resonance image K seed point nearest with it is carried out Fuzzy Correlation, and calculated every by 2-2
Fuzzy membership between a voxel and its K nearest seed：
Wherein,Indicate i-th of voxel in individual magnetic resonance image and seed point skBetween fuzzy membership, D (q, sk)
Indicate i-th of voxel in individual magnetic resonance image and seed point skThe distance between, m indicates the FUZZY WEIGHTED of fuzzy membership
Index, sk,st∈ S, S={ s1,s2,…,sK, S indicates the K nearest seed of i-th of voxel in individual magnetic resonance image
The set of point, k, t=1,2 ..., K；
2-3, the fuzzy membership calculated using step 2-2 are updated the seed point in individual magnetic resonance image space, obtain：
Wherein, NjIndicate the quantity with the voxel of j-th of seed point Fuzzy Correlation in individual magnetic resonance image space, ujrIndicate a
J-th of seed point in body magnetic resonance image space and the fuzzy membership with r-th of voxel of its Fuzzy Correlation, vr=[xr,yr,
zr,Ir]T, (xr,yr,zr) indicate seat with r-th of voxel of j-th of seed point Fuzzy Correlation in individual magnetic resonance image space
Mark, IrIndicate the gray value with r-th of voxel of j-th of seed point Fuzzy Correlation in individual magnetic resonance image space；
2-4 is carried out being iterated update to the seed point in individual magnetic resonance image space according to step 2-2 and 2-3, and is defined residual
Remaining error stops iteration when residual error is less than given threshold, and each super voxel is distributed to maximum fuzzy membership
Seed point to generate the super voxel of individual brain magnetic resonance image.
Step 3, after determining super voxel number, the super voxel of brain magnetic resonance image is generated by step 2.
2. a kind of super voxel generation method of the brain magnetic resonance image of iteration space fuzzy clustering according to claim 1,
It is characterized in that, the Gray homogeneity d between a-th of voxel and b-th of seed pointI(a, b)=| Ia-Ib|, wherein IbAnd IaRespectively
Indicate the image pixel intensities of b-th of seed point and a-th of voxel.
3. a kind of brain magnetic resonance image of iteration space fuzzy clustering according to claim 1
Super voxel generation method, which is characterized in that the space length between a-th of voxel and b-th of seed pointWherein (xb,yb,zb) and (xa,ya,za) b-th kind is indicated respectively
The coordinate of son point and a-th of voxel.
4. a kind of super voxel generation method of the brain magnetic resonance image of iteration space fuzzy clustering according to claim 1,
It is characterized in that, i-th of voxel in individual magnetic resonance image and seed point skThe distance between D (i, sk)=dI(i,sk)+λdS
(i,sk), dI(i,sk) indicate i-th of voxel in individual magnetic resonance image and seed point skBetween Gray homogeneity, dS(i,sk)
Indicate i-th of voxel in individual magnetic resonance image and seed point skBetween space length.
5. a kind of super voxel generation method of the brain magnetic resonance image of iteration space fuzzy clustering according to claim 1,
It is characterized in that, defined in step 2-4 update after and update before individual magnetic resonance image space seed point between it is European
Distance is residual error.
6. a kind of super voxel generation method of the brain magnetic resonance image of iteration space fuzzy clustering according to claim 1,
It is characterized in that, N=10000 in step 1-1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711445476.6A CN108305279B (en) | 2017-12-27 | 2017-12-27 | A kind of super voxel generation method of the brain magnetic resonance image of iteration space fuzzy clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711445476.6A CN108305279B (en) | 2017-12-27 | 2017-12-27 | A kind of super voxel generation method of the brain magnetic resonance image of iteration space fuzzy clustering |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108305279A true CN108305279A (en) | 2018-07-20 |
| CN108305279B CN108305279B (en) | 2019-07-05 |
Family
ID=62867799
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711445476.6A Active CN108305279B (en) | 2017-12-27 | 2017-12-27 | A kind of super voxel generation method of the brain magnetic resonance image of iteration space fuzzy clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108305279B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110473206A (en) * | 2019-07-24 | 2019-11-19 | 东南大学 | It is a kind of based on super voxel and the dispersion tensor image partition method for estimating study |
| CN110751664A (en) * | 2019-09-29 | 2020-02-04 | 东南大学 | Brain tissue segmentation method based on hyper-voxel matching |
| CN111081351A (en) * | 2019-12-02 | 2020-04-28 | 北京优脑银河科技有限公司 | Method and system for drawing brain function map |
| CN112580641A (en) * | 2020-11-23 | 2021-03-30 | 上海明略人工智能(集团)有限公司 | Image feature extraction method and device, storage medium and electronic equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103353987A (en) * | 2013-06-14 | 2013-10-16 | 山东大学 | Superpixel segmentation method based on fuzzy theory |
| US20150138191A1 (en) * | 2013-11-19 | 2015-05-21 | Thomson Licensing | Method and apparatus for generating superpixels |
| CN104809723A (en) * | 2015-04-13 | 2015-07-29 | 北京工业大学 | Three-dimensional liver CT (computed tomography) image automatically segmenting method based on hyper voxels and graph cut algorithm |
-
2017
- 2017-12-27 CN CN201711445476.6A patent/CN108305279B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103353987A (en) * | 2013-06-14 | 2013-10-16 | 山东大学 | Superpixel segmentation method based on fuzzy theory |
| US20150138191A1 (en) * | 2013-11-19 | 2015-05-21 | Thomson Licensing | Method and apparatus for generating superpixels |
| CN104809723A (en) * | 2015-04-13 | 2015-07-29 | 北京工业大学 | Three-dimensional liver CT (computed tomography) image automatically segmenting method based on hyper voxels and graph cut algorithm |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110473206A (en) * | 2019-07-24 | 2019-11-19 | 东南大学 | It is a kind of based on super voxel and the dispersion tensor image partition method for estimating study |
| CN110751664A (en) * | 2019-09-29 | 2020-02-04 | 东南大学 | Brain tissue segmentation method based on hyper-voxel matching |
| CN110751664B (en) * | 2019-09-29 | 2022-11-18 | 东南大学 | Brain tissue segmentation method based on hyper-voxel matching |
| CN111081351A (en) * | 2019-12-02 | 2020-04-28 | 北京优脑银河科技有限公司 | Method and system for drawing brain function map |
| CN111081351B (en) * | 2019-12-02 | 2024-02-20 | 北京优脑银河科技有限公司 | Brain functional map drawing method and system |
| CN112580641A (en) * | 2020-11-23 | 2021-03-30 | 上海明略人工智能(集团)有限公司 | Image feature extraction method and device, storage medium and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108305279B (en) | 2019-07-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | Brain tumor segmentation based on hybrid clustering and morphological operations | |
| Logeswari et al. | An improved implementation of brain tumor detection using segmentation based on soft computing | |
| CN108305279B (en) | A kind of super voxel generation method of the brain magnetic resonance image of iteration space fuzzy clustering | |
| US5933524A (en) | Method for segmentation of digital color images | |
| Wilson et al. | A class of discrete multiresolution random fields and its application to image segmentation | |
| Ortiz et al. | Improving MR brain image segmentation using self-organising maps and entropy-gradient clustering | |
| Higdon | Auxiliary variable methods for Markov chain Monte Carlo with applications | |
| Geva | Hierarchical unsupervised fuzzy clustering | |
| Logeswari et al. | An enhanced implementation of brain tumor detection using segmentation based on soft computing | |
| Bandhyopadhyay et al. | Segmentation of brain MRI image–a review | |
| Hu et al. | Texture segmentation based on a hierarchical Markov random field model | |
| Nie et al. | Two-dimensional extension of variance-based thresholding for image segmentation | |
| Schulz et al. | Non-linear hypothesis testing of geometric object properties of shapes applied to hippocampi | |
| Li et al. | Segmentation of pulmonary nodules using a GMM fuzzy C-means algorithm | |
| CN108364297A (en) | Blood-vessel image dividing method, terminal, storage medium | |
| Kirschner | The probabilistic active shape model: from model construction to flexible medical image segmentation | |
| Lu et al. | Textured image segmentation using autoregressive model and artificial neural network | |
| CN109285176B (en) | Brain tissue segmentation method based on regularization graph segmentation | |
| Nain et al. | Multiscale 3d shape analysis using spherical wavelets | |
| Afshar et al. | Lung tumor area recognition in CT images based on Gustafson-Kessel clustering | |
| CN111815640B (en) | Memristor-based RBF neural network medical image segmentation algorithm | |
| Roca et al. | Tractography-based parcellation of the cortex using a spatially-informed dimension reduction of the connectivity matrix | |
| Tansey et al. | False discovery rate smoothing | |
| Stough et al. | Regional appearance in deformable model segmentation | |
| CN116228753A (en) | Tumor prognosis evaluation method, device, computer equipment and storage medium |
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 |