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
- seed
- 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
- 210000004556 brain Anatomy 0.000 title claims abstract description 63
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000005070 sampling Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 abstract description 5
- 230000011218 segmentation Effects 0.000 description 5
- 210000001519 tissue Anatomy 0.000 description 5
- 238000010191 image analysis Methods 0.000 description 3
- 210000004885 white matter Anatomy 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000002490 cerebral effect Effects 0.000 description 2
- 210000001175 cerebrospinal fluid Anatomy 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 206010028980 Neoplasm Diseases 0.000 description 1
- 239000008186 active pharmaceutical agent Substances 0.000 description 1
- 230000004931 aggregating effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 210000004884 grey matter Anatomy 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
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 invention relates to a brain magnetic resonance image hyper-voxel generation method based on iterative spatial fuzzy clustering, and belongs to the field of digital images.
Background
The hyper-voxel technique is a process of aggregating voxels with highly redundant features into meaningful homogeneous regions. Compared with the conventional image processing basic unit voxel, the image analysis and processing based on a certain number of hyper-voxels can obtain better effect, and simultaneously, the efficiency can be greatly improved. The local region of the brain mr image is smoothed, which makes it suitable for hyper-voxel segmentation of the brain mr image. Recently, the hyper-voxel technique has been increasingly applied to brain magnetic resonance image analysis, and shows quite good performance in some applications, such as tumor localization and segmentation, tissue segmentation, image registration and functional grouping, etc. Therefore, a good hyper-voxel segmentation method of a brain magnetic resonance MRI image is crucial for the analysis of subsequent brain MRI images.
Brain MRI images have unique properties. First, the human brain has a complex internal structure, including several substructures of different sizes and morphologic complexity. The tissues in the brain can be generally divided into cerebrospinal fluid, white matter and white matter, which have complex boundary shapes and topologies. The conventional method of generating seed points by uniform sampling in a lattice structure may generate voxels of inhomogeneous brain mr images. Secondly, due to the limitation of insufficient resolution in the magnetic resonance image, the loss of contrast between adjacent tissues results in the possibility of multiple tissues being contained within a voxel, i.e. partial volume effects. Conventional hyper-voxel generation algorithms for natural images employ an otherwise-rigid clustering method to cluster voxels. The super-voxel segmentation of the brain magnetic resonance image by using a hard clustering algorithm has the result that a plurality of tissues (including gray matter, white matter and cerebrospinal fluid) are contained in one super-voxel, so that the boundary fitting capability of the super-voxel is very poor.
Currently, there are hyper-voxel methods developed for natural images, such as simple linear iterative methods, graph-based methods, mean shift, etc. hyper-voxel generation algorithms, and applied in many fields of computer vision, with reasonably good results. However, generating suitable hyper-voxels for a cerebral magnetic resonance image remains challenging in cerebral magnetic resonance image analysis. Generating brain MRI image hyper-voxels purely with existing algorithms is clearly not suitable because these algorithms ignore the special properties of brain magnetic resonance images for natural images.
Disclosure of Invention
The invention provides a super voxel generation method of a brain magnetic resonance image based on iterative spatial fuzzy clustering, which aims at generating ideal super voxels of the brain magnetic resonance image.
The invention adopts the following technical scheme for solving the technical problems:
the invention provides a brain magnetic resonance image hyper-voxel generation method of iterative spatial fuzzy clustering, which comprises the following steps:
step 1, constructing a seed template from a population-based brain magnetic resonance template image, and then projecting the seed template to an individual space to generate seed points of the individual space, wherein the method specifically comprises the following steps:
step 1, constructing a seed template from a population-based brain magnetic resonance template image, and then projecting the seed template to an individual space to generate seed points of the individual space, wherein the method specifically comprises the following steps:
1-1, uniformly sampling N seed points in a brain region of a brain magnetic resonance template image;
1-2, calculating the distance between each voxel on the magnetic resonance template image of the brain and the seed point of the step 1-1, wherein the distance D (a, b) between the a-th voxel and the b-th seed point is DI(a,b)+λdS(a,b)，dI(a, b) is the gray scale distance between the a-th voxel and the b-th seed point, dS(a, b) is the spatial distance between the a-th voxel and the b-th seed point, and lambda is the weight of the spatial distance and the gray scale distance;
1-3, merging two seed points with the minimum distance into a new seed point each time by adopting a hierarchical clustering method to generate a seed point template;
1-4, projecting a corresponding seed point template to the space of the individual magnetic resonance image in a mode of registering the brain magnetic resonance template image to the individual magnetic resonance image to generate seed points in the space of the individual magnetic resonance image;
step 2, calculating fuzzy membership degrees between voxels of the individual magnetic resonance image and the nearest K seeds by adopting an iterative spatial fuzzy clustering method to generate the hyper-voxels, wherein the method specifically comprises the following steps:
2-1, initializing the seed point of the individual magnetic resonance image space in the step 1-4 by using a vector, wherein the initialization vector of the jth seed point of the individual magnetic resonance image space(xj,yj,zj) Is the coordinate of the jth seed point of the individual magnetic resonance image space,is the j seed point in the individual magnetic resonance image space and the volume of all voxels in its 3 x 3 neighborhoodAverage of the intensity of the elements;
2-2, carrying out fuzzy association on each voxel in the individual magnetic resonance image and the nearest K seed points thereof, and calculating fuzzy membership between each voxel and the nearest K seeds:
wherein,representing the ith voxel and seed point s in an individual magnetic resonance imagekFuzzy degree of membership between, D (q, s)k) Representing the ith voxel and seed point s in an individual magnetic resonance imagekM represents a fuzzy weight index of fuzzy membership, sk,st∈S，S＝{s1,s2,…,sKS denotes the set of the nearest K seed points for the ith voxel in the individual magnetic resonance image, K, t-1, 2, …, K;
2-3, updating the seed points of the individual magnetic resonance image space by using the fuzzy membership degree calculated in the step 2-2 to obtain:
wherein N isjRepresenting the number of voxels, u, associated with the jth seed point blur of the individual magnetic resonance image spacejrRepresenting the fuzzy membership, v, of the jth seed point of the individual magnetic resonance image space and the r-th voxel in fuzzy association therewithr＝[xr,yr,zr,Ir]T，(xr,yr,zr) Representing the coordinates of the r-th voxel associated with the j-th seed point blur of the individual magnetic resonance image space, IrRepresenting the r-th voxel associated with the j-th seed point blur of the individual magnetic resonance image spaceGray value;
and 2-4, iteratively updating the seed points of the individual magnetic resonance image space according to the steps 2-2 and 2-3, defining a residual error, stopping iteration when the residual error is smaller than a set threshold value, and distributing each hyper-voxel to the seed point with the maximum fuzzy membership degree to generate the hyper-voxel of the individual brain magnetic resonance image.
And 3, after the number of the super voxels is determined, generating the super voxels of the brain magnetic resonance image through the step 2.
As a further technical scheme of the invention, the gray scale distance d between the a-th voxel and the b-th seed pointI(a,b)＝|Ia-IbI, wherein IbAnd IaRepresenting the pixel intensities of the b-th seed point and the a-th voxel, respectively.
As a further technical scheme of the invention, the spatial distance between the a-th voxel and the b-th seed pointWherein (x)b,yb,zb) And (x)a,ya,za) Respectively representing the coordinates of the b-th seed point and the a-th voxel.
As a further technical scheme of the invention, the ith voxel and the seed point s in the individual magnetic resonance imagekA distance D (i, s) therebetweenk)＝dI(i,sk)+λdS(i,sk)，dI(i,sk) Representing the ith voxel and seed point s in an individual magnetic resonance imagekGray scale distance between, dS(i,sk) Representing the ith voxel and seed point s in an individual magnetic resonance imagekThe spatial distance therebetween.
As a further technical solution of the present invention, the euclidean distance between the seed points of the individual magnetic resonance image space after the update and before the update is defined as a residual error in step 2-4.
As a further embodiment of the present invention, N in step 1-1 is 10000.
Compared with the prior art, the invention adopting the technical scheme has the following technical effects: the invention discloses a brain magnetic resonance image hyper-voxel generation method of iterative spatial fuzzy clustering, which comprises the following steps that firstly, a group of seed templates are obtained from a brain MRI template based on a population because human brains have the same topological structure; secondly, in order to eliminate the influence of partial volume effect, an iterative spatial fuzzy clustering algorithm is provided, and voxels are distributed to each seed to generate the hyper-voxels. The invention can be better applied to the magnetic resonance image of the brain to generate the effective hyper-voxel of the magnetic resonance image of the brain.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
FIG. 2 is a schematic flow chart of the present invention.
Fig. 3 is a brain magnetic resonance image.
Fig. 4 shows the results of the super-voxel generation in the method according to the present invention, where (a) shows the result of 500 super-voxels, (b) shows the result of 1000 super-voxels, and (c) shows the result of 2000 super-voxels.
Detailed Description
The technical scheme of the invention is further explained in detail by combining the attached drawings:
the invention provides a brain magnetic resonance image hyper-voxel generation method of iterative spatial fuzzy clustering, as shown in figures 1 and 2, firstly, a group of seed templates are constructed from a population-based brain magnetic resonance image template; then, calculating fuzzy membership degrees between the voxels and the nearest K seeds by adopting an iterative spatial fuzzy clustering method to generate the hyper-voxels; finally, after determining the number of voxels, the voxels are generated according to step 1 and step 2.
Firstly, the generation of the seed template specifically comprises the following steps: a seed template is constructed from population-based brain MRI templates and then projected into individual space to produce reliable seeds in hyper-voxel generation, comprising the following 4 steps:
(1-1) uniformly sampling N seed points in a brain region of a brain magnetic resonance template image.
(1-2) calculating the space distance d between each voxel on the magnetic resonance template image of the brain and the seed point in the step 1-1SDistance d from gray levelIWherein d isI、dSThe calculation formula of (a) is as follows:
dI(a,b)＝|Ia-Ib| (1)
in the above formula, dI(a, b) is the gray scale distance between the a-th voxel and the b-th seed point, IbAnd IaRepresenting the pixel intensities of the b-th seed point and the a-th voxel, respectively.
In the above formula, dS(a, b) is the spatial distance between the a-th voxel and the b-th seed point, (x)b,yb,zb) And (x)a,ya,za) Respectively representing the coordinates of the b-th seed point and the a-th voxel.
By a spatial distance dSDistance d from gray levelICalculating the distance D:
D(a,b)＝dI(a,b)+λdS(a,b) (3)
in the above formula, D (a, b) is the distance between the a-th voxel and the b-th seed point, and λ represents the spatial distance DSDistance d from gray levelIThe weight of (c).
(1-3) generating a seed point template by merging two seeds having the minimum distance each time into a new seed using the prior-disclosed hierarchical clustering method based on the distance D calculated in the step 1-2. A weighted combination of the spatial distance and voxel intensity of neighboring seeds will be used as a distance measure between seeds in merging the seeds.
And (1-4) projecting the corresponding seed template to the space of the individual magnetic resonance image by registering the brain magnetic resonance template image to the individual magnetic resonance image to generate the seed points of the space of the individual magnetic resonance image.
Then, the iterative spatial fuzzy clustering specifically operates as:
(2-1) initializing each seed point in step 1-4 with a vector, the initialization vector C of the jth seed point of the individual magnetic resonance image spacejIs defined as follows:
wherein (x)j,yj,zj) Is the coordinate of the jth seed point of the individual magnetic resonance image space;the average value of the voxel intensities of the jth seed point in the individual magnetic resonance image space and all voxels in the 3 x 3 neighborhood thereof can inhibit the influence of noise and improve the robustness of the algorithm.
(2-2) fuzzy associating each voxel in the individual magnetic resonance image with its nearest K seeds. And calculating fuzzy membership between each voxel and the nearest K seeds, wherein the fuzzy membership calculation formula is as follows:
wherein,representing the ith voxel and seed point s in an individual magnetic resonance imagekM represents a fuzzy weighted index of fuzzy membership, sk,st∈S，S＝{s1,s2,…,sKS denotes the set of the nearest K seed points for the ith voxel in the individual magnetic resonance image, K, t 1,2, …, K.
D(i,sk) Representing the ith voxel and seed point s in an individual magnetic resonance imagekThe distance between, is defined as in step 1-2:
D(i,sk)＝dI(i,sk)+λdS(i,sk) (6)
wherein d isI(i,sk) Representing the ith voxel and seed point s in an individual magnetic resonance imagekGray scale distance between, dS(i,sk) Representing the ith voxel and seed point s in an individual magnetic resonance imagekThe spatial distance therebetween.
(2-3) updating the seed points by using the fuzzy membership degree calculated in the step 2-2, wherein an updating formula is as follows:
wherein N isjIs a function representing the number of voxels, u, associated with the jth seed point blur of the individual magnetic resonance image spacejrRepresenting the fuzzy membership, v, of the jth seed point of the individual magnetic resonance image space and the r-th voxel in fuzzy association therewithrIs a vector description of the r-th voxel (similar to the vector description of the seed) associated with the j-th seed point blur of the individual magnetic resonance image space, represented by:
vr＝[xr,yr,zr,Ir]T(8)
wherein (x)r,yr,zr) Representing the coordinates of the r-th voxel associated with the j-th seed point blur of the individual magnetic resonance image space, IrRepresenting the gray value of the r-th voxel associated with the j-th seed point blur of the individual magnetic resonance image space.
And (2-4) iteratively updating the seed points of the individual magnetic resonance image space according to the steps 2-2 and 2-3, defining a residual error, stopping iteration when the residual error is smaller than a set threshold value, and distributing each super voxel to the seed point with the maximum fuzzy membership degree to generate the super voxel of the individual brain magnetic resonance image.
Finally, after determining the number of voxels, seed points of the individual space are generated by step 1, and voxels of the magnetic resonance image of the brain are generated by step 2.
In the invention, a large number of seeds are uniformly sampled in the brain area of the brain image, and the iterative spatial fuzzy clustering algorithm of the invention is used for generating the seeds. These obtained seeds are then further combined by a hierarchical clustering algorithm that preserves the spatial contiguity of the clusters to generate the seed template. In merging the seeds, the two seeds with the smallest distance are merged into one new seed using a weighted combination of the spatial distance and voxel intensity of the neighboring seeds as a distance measure between the seeds. The corresponding seed template is projected into the individual space by means of registering the brain template magnetic resonance image to the individual image. The projection is performed by pair-wise registering the template image with the single image using an efficient Elastix tool.
Example (b):
the brain magnetic resonance hyper-voxel generation algorithm for iterative spatially fuzzy clustering of the present invention is described below using the brain web18 dataset data as an example.
The experimental conditions are as follows: a computer equipped with an Intel processor (3.4GHz) and 10GB RAM, 64-bit operating system, and Matlab (version R2014 a) programming language was selected for the experiments.
The experimental data were brain magnetic resonance images of the brain web18 dataset. Each MRI image comprises 181 × 217 × 181 voxels with a size of 1mm × 1mm × 1 mm. The experimental parameters were set as: k is 6, λ is 0.8, and the number of superpixels is set to 500,1000, and 2000, respectively. Fig. 3 shows a brain MRI image original graph of the brain web18, in fig. 4, (a) shows a result of 500 number of voxels, in fig. 4, (b) shows a result of 1000 number of voxels, and in fig. 4, (c) shows a result of 2000 number of voxels.
The above description is only an embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can understand that the modifications or substitutions within the technical scope of the present invention are included in the scope of the present invention, and therefore, the scope of the present invention should be subject to the protection scope of the claims.
Claims (6)
1. A brain magnetic resonance image hyper-voxel generation method of iterative spatial fuzzy clustering is characterized by comprising the following steps:
step 1, constructing a seed template from a population-based brain magnetic resonance template image, and then projecting the seed template to an individual space to generate seed points of the individual space, wherein the method specifically comprises the following steps:
1-1, uniformly sampling N seed points in a brain region of a brain magnetic resonance template image;
1-2, calculating the distance between each voxel on the magnetic resonance template image of the brain and the seed point of the step 1-1, and obtaining the distanceIn (2), the distance D (a, b) ═ D between the a-th voxel and the b-th seed pointI(a,b)+λdS(a,b)，dI(a, b) is the gray scale distance between the a-th voxel and the b-th seed point, dS(a, b) is the spatial distance between the a-th voxel and the b-th seed point, and lambda is the weight of the spatial distance and the gray scale distance;
1-3, merging two seed points with the minimum distance into a new seed point each time by adopting a hierarchical clustering method to generate a seed point template;
1-4, projecting a corresponding seed point template to the space of the individual magnetic resonance image in a mode of registering the brain magnetic resonance template image to the individual magnetic resonance image to generate seed points in the space of the individual magnetic resonance image;
step 2, calculating fuzzy membership degrees between voxels of the individual magnetic resonance image and the nearest K seeds by adopting an iterative spatial fuzzy clustering method to generate the hyper-voxels, wherein the method specifically comprises the following steps:
2-1, initializing the seed point of the individual magnetic resonance image space in the step 1-4 by using a vector, wherein the initialization vector of the jth seed point of the individual magnetic resonance image space(xj,yj,zj) Is the coordinate of the jth seed point of the individual magnetic resonance image space,is the average of the voxel intensities of the jth seed point in the individual magnetic resonance image space and all voxels in its 3 x 3 neighborhood;
2-2, carrying out fuzzy association on each voxel in the individual magnetic resonance image and the nearest K seed points thereof, and calculating fuzzy membership between each voxel and the nearest K seeds:
wherein,representing the ith voxel and seed point s in an individual magnetic resonance imagekFuzzy degree of membership between, D (q, s)k) Representing the ith voxel and seed point s in an individual magnetic resonance imagekM represents a fuzzy weight index of fuzzy membership, sk,st∈S，S＝{s1,s2,…,sKS denotes the set of the nearest K seed points for the ith voxel in the individual magnetic resonance image, K, t-1, 2, …, K;
2-3, updating the seed points of the individual magnetic resonance image space by using the fuzzy membership degree calculated in the step 2-2 to obtain:
wherein N isjRepresenting the number of voxels, u, associated with the jth seed point blur of the individual magnetic resonance image spacejrRepresenting the fuzzy membership, v, of the jth seed point of the individual magnetic resonance image space and the r-th voxel in fuzzy association therewithr＝[xr,yr,zr,Ir]T，(xr,yr,zr) Representing the coordinates of the r-th voxel associated with the j-th seed point blur of the individual magnetic resonance image space, IrRepresenting a gray value of an r-th voxel associated with a j-th seed point blur of the individual magnetic resonance image space;
and 2-4, iteratively updating the seed points of the individual magnetic resonance image space according to the steps 2-2 and 2-3, defining a residual error, stopping iteration when the residual error is smaller than a set threshold value, and distributing each hyper-voxel to the seed point with the maximum fuzzy membership degree to generate the hyper-voxel of the individual brain magnetic resonance image.
And 3, after the number of the super voxels is determined, generating the super voxels of the brain magnetic resonance image through the step 2.
2. According toThe method of claim 1, wherein the distance d between the a-th voxel and the b-th seed point is a gray scale distanceI(a,b)＝|Ia-IbI, wherein IbAnd IaRepresenting the pixel intensities of the b-th seed point and the a-th voxel, respectively.
3. The method of claim 1, wherein the spatial distance between the a-th voxel and the b-th seed point is a spatial distanceWherein (x)b,yb,zb) And (x)a,ya,za) Respectively representing the coordinates of the b-th seed point and the a-th voxel.
4. The method of claim 1, wherein the i-th voxel and the seed point s in the individual magnetic resonance image are selected from the group consisting ofkA distance D (i, s) therebetweenk)＝dI(i,sk)+λdS(i,sk)，dI(i,sk) Representing the ith voxel and seed point s in an individual magnetic resonance imagekGray scale distance between, dS(i,sk) Representing the ith voxel and seed point s in an individual magnetic resonance imagekThe spatial distance therebetween.
5. The method of claim 1, wherein the Euclidean distance between seed points of the individual MR image space after and before updating is defined as the residual error in step 2-4.
6. The method for generating hypervoxels in magnetic resonance images of brain according to claim 1, wherein N is 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 (5)
| 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 |
| CN114581451A (en) * | 2022-02-24 | 2022-06-03 | 东南大学 | Scattering map neural network-based brain magnetic resonance image segmentation method |
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 (8)
| 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 |
| CN112580641B (en) * | 2020-11-23 | 2024-06-04 | 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 |
| CN114581451A (en) * | 2022-02-24 | 2022-06-03 | 东南大学 | Scattering map neural network-based brain magnetic resonance image segmentation method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108305279B (en) | 2019-07-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108305279B (en) | A kind of super voxel generation method of the brain magnetic resonance image of iteration space fuzzy clustering | |
| Lian et al. | Joint tumor segmentation in PET-CT images using co-clustering and fusion based on belief functions | |
| Khosravanian et al. | Fast level set method for glioma brain tumor segmentation based on Superpixel fuzzy clustering and lattice Boltzmann method | |
| Zhang et al. | Brain tumor segmentation based on hybrid clustering and morphological operations | |
| Zhang et al. | Review of breast cancer pathologigcal image processing | |
| Ma et al. | Automated nasopharyngeal carcinoma segmentation in magnetic resonance images by combination of convolutional neural networks and graph cut | |
| Chen et al. | Automatic X-ray landmark detection and shape segmentation via data-driven joint estimation of image displacements | |
| CN104933709B (en) | Random walk CT lung tissue image automatic segmentation methods based on prior information | |
| Zhong et al. | Boosting‐based cascaded convolutional neural networks for the segmentation of CT organs‐at‐risk in nasopharyngeal carcinoma | |
| Göçeri | Fully automated liver segmentation using Sobolev gradient‐based level set evolution | |
| Schulz et al. | Non-linear hypothesis testing of geometric object properties of shapes applied to hippocampi | |
| Roy et al. | International journal of advanced research in computer science and software engineering | |
| He et al. | MetricUNet: Synergistic image-and voxel-level learning for precise prostate segmentation via online sampling | |
| Li et al. | Segmentation of pulmonary nodules using a GMM fuzzy C-means algorithm | |
| Peng et al. | H-SegMed: a hybrid method for prostate segmentation in TRUS images via improved closed principal curve and improved enhanced machine learning | |
| Kong et al. | Iterative spatial fuzzy clustering for 3D brain magnetic resonance image supervoxel segmentation | |
| Kumar et al. | Lung nodule segmentation using 3-dimensional convolutional neural networks | |
| Jaffar et al. | An ensemble shape gradient features descriptor based nodule detection paradigm: a novel model to augment complex diagnostic decisions assistance | |
| Chung et al. | Statistical analysis on brain surfaces | |
| Banerjee et al. | A CADe system for gliomas in brain MRI using convolutional neural networks | |
| Zheng et al. | A variational approach to liver segmentation using statistics from multiple sources | |
| Afshar et al. | Lung tumor area recognition in CT images based on Gustafson-Kessel clustering | |
| CN109285176B (en) | Brain tissue segmentation method based on regularization graph segmentation | |
| Zhang et al. | Topology-preserving segmentation network: A deep learning segmentation framework for connected component | |
| Yamasaki et al. | GrowCut-based fast tumor segmentation for 3D magnetic resonance images |
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 |