CN108898602A - A kind of FCM medical image cutting method based on improvement QPSO - Google Patents
A kind of FCM medical image cutting method based on improvement QPSO Download PDFInfo
- Publication number
- CN108898602A CN108898602A CN201810677834.4A CN201810677834A CN108898602A CN 108898602 A CN108898602 A CN 108898602A CN 201810677834 A CN201810677834 A CN 201810677834A CN 108898602 A CN108898602 A CN 108898602A
- Authority
- CN
- China
- Prior art keywords
- medical image
- particle
- pixel
- fcm
- cluster centre
- 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.)
- Pending
Links
Classifications
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/03—Recognition of patterns in medical or anatomical images
Abstract
The present invention is directed in medical image between different soft tissues or between soft tissue and lesion the features such as obscurity boundary, fine structure complex distribution, introduce Fuzzy clustering techniques, and optimized using improved quanta particle swarm optimization, a kind of image segmentation scheme is proposed on this basis.The present invention effectively improves standard fuzzy C-mean algorithm fuzzy clustering algorithm and relies on initial cluster center, be easy to fall into the defect of local optimum, so that medical image is preferably divided using a kind of new improvement quantum particle swarm optimization.Since convergence can be effectively performed in method provided by the invention after given primary condition, this method is for having preferable effect the problems such as ever-present fuzzy and obscure boundary in processing medical image.More raw informations can be retained during handling medical image using method provided by the invention, robustness is higher than other partitioning algorithms such as hardness cluster.
Description
Technical field
The invention discloses a kind of FCM medical image cutting method based on QPSO, belongs to technical field of image processing, main
It is used for the processing of medical image.
Background technique
Medical imaging instrument provides image information abundant for medical diagnosis, it include many useful medicals diagnosis on disease and
Analyze information.These medical images are efficiently used, doctor can effectively be helped to carry out computer-aided diagnosis, implement intervention
Formula treatment formulates internal surgical procedures planning, carries out dynamic analog to corresponding medical tissue organ and analyze the knot of diseased region
Structure and generating process improve the accuracy rate of medical diagnosis on disease.
Under normal conditions, medical image is extremely complex.This is because the complexity of human anatomic structure itself and
Inevitably by noise, field offset effect, local bulk effect etc. during obtaining image using various image technologies
Influence so that the image got often appear as it is noisy it is high, contrast is low, organization internal gray scale is uneven, soft group different
Between knitting or between soft tissue and lesion the features such as obscurity boundary, fine structure complex distribution.Such as medically common MR
(Magnetic Resonance) brain image, generally includes cortex, subcutaneous fat, skull, white matter of brain, grey matter, cerebrospinal fluid
Equal Various Tissues, usually mutual aliasing (is covered with grey matter outside white matter, grey matter between every kind of tissue profile complexity, different tissues
In be filled with cerebrospinal fluid), without clearly boundary.It just needs to handle image when this.
Medical Image Processing is the first step of medical imaging analysis, helps to keep image more intuitive, clear, raising is examined
Disconnected efficiency, all attaches great importance to medical image processing both at home and abroad.And image segmentation is committed step of the image procossing to image analysis.
Image segmentation is a kind of technology in region for dividing the image into several homogeneous areas and extracting complex scene needs, according to some equal
Even property standard, such as color, intensity or texture, region is similar, thus the boundary in positioning and identification segmented image.Existing image
Dividing method mainly divides following a few classes:Dividing method based on threshold value, the dividing method based on region, the segmentation side based on edge
Method and the dividing method based on specific theory etc..However segmentation precision is lower on the whole for classical image segmentation algorithm, and
Big multipair noise-sensitive, it is difficult to obtain large-scale application.
The characteristics of medical image, brings sizable difficulty to the research of image Segmentation Technology, while also giving fuzzy clustering
Technology FCM brings ample scope for abilities.Fuzzy clustering techniques describe complication system in the way of less accurate, can be effectively
Ambiguity present in medical image is portrayed, there is very strong specific aim to the image as this kind of obscurity boundary of MR brain image.
The partitioning algorithm that these features of fuzzy clustering algorithm keep it more traditional than other in terms of handling medical image segmentation has more advantage,
Therefore it is widely studied and applied.Clustering has unsupervised property, high efficiency compared with traditional medicine image partition method
And adaptivity, there is preferable segmentation accuracy rate to the higher image of quality.But there is also following defects for such methods:For
Unknown complicated medical image is difficult to judge that it should be divided into several classes and FCM clustering algorithm to initial cluster center
It is stronger with subordinated-degree matrix dependence, it is easy to fall into locally optimal solution.The present invention is directed to disadvantages described above, using grey level histogram and
Improved quanta particle swarm optimization optimizes, and can effectively improve the quality of medical image segmentation.
Summary of the invention
The present invention is directed in medical image between different soft tissues or obscurity boundary, subtle knot between soft tissue and lesion
The features such as structure complex distribution, is introduced Fuzzy clustering techniques, and is optimized using improved quanta particle swarm optimization, basic herein
On propose a kind of image segmentation scheme.
The application proposes a kind of FCM medical image cutting method based on improvement QPSO, comprises the steps of:
Step 1:Input a medical image;
Step 2:Fuzzy Exponential m, iteration stopping threshold value e are set, and gray feature is extracted to given medical image, does ash
Histogram is spent, determines that clusters number is c;
Step 3:Optimal cluster center is found using improved quantum particle swarm;
Step 4:Using obtained best particle, the i.e. cluster centre of medical image, each pixel in medical image is calculated
The degree of membership of point to cluster centre determines the last ownership of each pixel in medical image using it as foundation;
Step 5:Export final medical image segmentation result.
Further, in the step 3 using improved quantum particle swarm find Optimal cluster center the specific steps are:
Step 3.1:Parameter initialization, taking c pixel at random as cluster centre is a primary, is repeated
N times, symbiosis is at n primary;
Step 3.2:Degree of membership is calculated using Fuzzy Exponential m and cluster centre, using the objective function of FCM as improved amount
The fitness function of seed swarm optimization calculates fitness value;
Step 3.3:The fitness value for comparing each particle carries out fitness evaluation to all particles.Utilize improved amount
Seed subgroup more new particle, and its fitness value is calculated, if reaching termination condition, stop iteration, otherwise continues to update grain
Son.
Further, the specific calculating step of the step 3.1 is:Parameter initialization, it is assumed that X=x1, x2 ... xt}
A series of pixels are represented, t is the number of pixel in medical image, Population Size n, maximum number of iterations M is arranged, wherein 2<c<
t;It is a primary that c pixel is taken at random as cluster centre, n times is repeated, symbiosis is at n primary, often
A kind of cluster centre can be expressed as V={ v1, v2... vc}。
Further, the specific calculating step of the step 3.2 is:Each pixel is calculated into cluster according to formula (1)
The distance d of the heartij, the degree of membership u that each pixel corresponds to each cluster centre is calculated according to distance and formula (2)ij, wherein m is to be subordinate to
Category degree index, portrays the fog-level of classification results, m>1, m bigger presentation class result is fuzzyyer, usually takes m=2;Formula (3)
Middle J (U, V) indicates the fitness value of particle, its size reflection image pixel is worth smaller expression to the subjection degree of target class
Pixel belongs to that a possibility that corresponding class is bigger, and Clustering Effect is better, while to meet formula (4).
dij=| | xi-vj||2 (1)
Further, the specific calculating step of the step 3.3 is:Formula such as formula (5) to (9) institute that particle position updates
Show, wherein pbiIt is the personal best particle of i-th of particle;Gb is the optimal location of entire population;Rand (0,1) is a letter
Number, return value is the random number between one [0,1];T is current evolutionary generation;tmaxIt is the maximum evolutionary generation of algorithm;
β is known as shrinkage expansion coefficient, it is gradually reduced with the iteration of algorithm;piIt is the position of i-th of particle.Utilize what is newly obtained
Particle calculates new fitness value, if reached, whether termination condition i.e. fitness value is less than threshold value e or the number of iterations is more than
Otherwise M continues more new particle.
A=rand (0,1) (5)
P=a*pbi+(1-a)*gb (6)
U=rand (0,1) (8)
The beneficial effect that the present invention reaches:
(1) a kind of new improvement quantum particle swarm is utilized for the deficiency of medical image segmentation for fuzzy C-means clustering
Optimization algorithm searches for global optimum's cluster centre, enhances ability of searching optimum and precision, effectively improves standard fuzzy C-mean algorithm mould
It pastes clustering algorithm to rely on initial cluster center, is easy to fall into the defect of local optimum, so that medical image obtains more preferably
Segmentation.
(2) for unknown complicated medical image, the number of typical peak is obtained using its grey level histogram
As the number c of cluster centre, grey-level statistics are often simple and effective, and this information is language intrinsic in image
Method information does not change with the viewing angle of observer.
(3) convergence, this method can be effectively performed after given primary condition due to method provided by the invention
For there is preferable effect the problems such as ever-present fuzzy and obscure boundary in processing medical image.It is mentioned using the present invention
The method of confession can retain more raw informations during handling medical image, and robustness is higher than hardness cluster etc.
Other partitioning algorithms.
Detailed description of the invention
Fig. 1 is medical image segmentation broad flow diagram of the present invention.
Fig. 2 is the specific flow chart that Optimal cluster center is found using improved quantum particle swarm.
Specific embodiment
Technical solution of the present invention is described in further detail with reference to the accompanying drawings of the specification.
The present invention is easy to fall into locally optimal solution etc. for unknown complicated medical image and FCM clustering algorithm and lacks
It falls into, makes improvements.Clusters number is judged first with the grey level histogram of image, is replaced using improved QPSO algorithm
The Gradient Iteration process of FCM finds out Optimal cluster centers and calculates degree of membership, is that foundation is split medical image with it.
The algorithm reduces dependence of the FCM algorithm to initial cluster center, while enhancing ability of searching optimum, accelerates convergence rate.It should
The process of scheme is as shown in Figure 1, be below described in further details the present invention:
Step 1;Input a medical image.
Step 2:Fuzzy Exponential m, iteration stopping threshold value e are set, and gray feature is extracted to given medical image, does ash
Histogram is spent, determines that clusters number is c.
Step 3:Optimal cluster center is found using improved quantum particle swarm, flow chart is shown in Fig. 2.
Step 3.1:Assuming that X={ x1, x2... xtA series of pixels are represented, t is the number of pixel in medical image,
Population Size n, maximum number of iterations M are set, wherein 2<c<t.It is an initial grain that c pixel is taken at random as cluster centre
Son, is repeated n times, and symbiosis can be expressed as V={ v at n primary, the cluster centre of every one kind1, v2... vc}。
Step 3.2:According to the distance d of each pixel of formula (1) calculating to cluster centreij, calculated according to distance and formula (2)
Each pixel corresponds to the degree of membership u of each cluster centreij, wherein m is degree of membership index, portrays the fuzzy journey of classification results
Degree, m>1, m bigger presentation class result is fuzzyyer, usually takes m=2.J (U, V) indicates the fitness value of particle in formula (3), it
Size reflection image pixel to the subjection degree of target class, a possibility that smaller expression pixel of value belongs to corresponding class, is bigger, poly-
Class effect is better, while to meet formula (4).
dij=| | xi-vj||2 (1)
Step 3.3:In group, due to initializing and updating the randomness of iterative process, the initial position of each particle,
Evolution process differs greatly, and individual optimal value pbest difference is also larger, if simply putting down all positions pbest progress arithmetic
, the influence power of outstanding particle can be reduced, the reduction of ability and convergence rate is solved so as to cause algorithm.This is changed
Into, after carrying out fitness evaluation to all particles, retain the optimal value pbest and group optimal value gbest of each particle, and
All particles are ranked up by pbest, to the particle after sequence, member of the particle of k ratio as outstanding particle group before taking,
Take a particle as outstanding particle (ymb) at random again.The value of k according to problem to be solved, population scale and can be commented by user
The factors such as valence function determine.On this basis, the formula such as formula (5) that particle position updates -- shown in (9), wherein pbiIt is i-th
The personal best particle of a particle;Gb is the optimal location of entire population;Rand (0,1) is a function, and return value is one
Random number between a [0,1];T is current evolutionary generation;tmaxIt is the maximum evolutionary generation of algorithm;β is known as shrinkage expansion system
Number, it is gradually reduced with the iteration of algorithm;piIt is the position of i-th of particle.New fit is calculated using the particle newly obtained
Angle value is answered, if reaching termination condition i.e. fitness value whether to be less than threshold value e or the number of iterations is more than M, otherwise continues to update
Particle.
A=rand (0,1) (5)
P=a*pbi+(1-a)*gb (6)
U=rand (0,1) (8)
Step 4:Using obtained best particle, the i.e. cluster centre of medical image, each pixel in medical image is calculated
The degree of membership of point to cluster centre determines the last ownership of each pixel in medical image using it as foundation.
Step 5:Export final medical image segmentation result.
The foregoing is merely better embodiment of the invention, protection scope of the present invention is not with above embodiment
Limit, as long as those of ordinary skill in the art's equivalent modification or variation made by disclosure according to the present invention, should all be included in power
In the protection scope recorded in sharp claim.
Claims (5)
1. one kind is based on FCM (the Fuzzy C- for improving QPSO (Quantum Particle Swarm Optimization)
Means) medical image cutting method, it is characterised in that comprise the steps of:
Step 1:Input a medical image;
Step 2:Fuzzy Exponential m, iteration stopping threshold value e are set, and gray feature is extracted to given medical image, it is straight to do gray scale
Fang Tu determines that clusters number is c;
Step 3:Optimal cluster center is found using improved quantum particle swarm;
Step 4:Using obtained best particle, the i.e. cluster centre of medical image, calculates each pixel in medical image and arrive
The degree of membership of cluster centre determines the last ownership of each pixel in medical image using it as foundation;
Step 5:Export final medical image segmentation result.
2. according to claim 1 a kind of based on the FCM medical image cutting method for improving QPSO, it is characterised in that:Institute
State in step 3 using improved quantum particle swarm find Optimal cluster center the specific steps are:
Step 3.1:Parameter initialization, taking c pixel at random as cluster centre is a primary, and n times are repeated,
Symbiosis is at n primary;
Step 3.2:Degree of membership is calculated using Fuzzy Exponential m and cluster centre, using the objective function of FCM as improved quantum grain
The fitness function of swarm optimization calculates fitness value;
Step 3.3:The fitness value for comparing each particle carries out fitness evaluation to all particles.Utilize improved quantum grain
Subgroup more new particle, and its fitness value is calculated, if reaching termination condition, stop iteration, otherwise continues more new particle.
3. according to claim 2 a kind of based on the FCM medical image cutting method for improving QPSO, it is characterised in that:Institute
The specific calculating step for stating step 3.1 is:Parameter initialization, it is assumed that X={ x1, x2... xtA series of pixels are represented, t is
Population Size n, maximum number of iterations M is arranged, wherein 2 in the number of pixel in medical image<c<t;C pixel is taken to make at random
It is a primary for cluster centre, n times is repeated, at n primary, the cluster centre of every one kind can be with table for symbiosis
It is shown as V={ v1, v2... vc}。
4. according to claim 2 a kind of based on the FCM medical image cutting method for improving QPSO, it is characterised in that:Institute
The specific calculating step for stating step 3.2 is:According to the distance d of each pixel of formula (1) calculating to cluster centreij, according to distance
And formula (2) calculates the degree of membership u that each pixel corresponds to each cluster centreij, wherein m is degree of membership index, portrays classification knot
The fog-level of fruit, m>1, m bigger presentation class result is fuzzyyer, usually takes m=2;J (U, V) indicates the suitable of particle in formula (3)
Angle value is answered, subjection degree of its size reflection image pixel to target class, the smaller possibility for indicating pixel and belonging to corresponding class of value
Property is bigger, and Clustering Effect is better, while to meet formula (4).
dij=| | xi-vj||2 (1)
5. according to claim 2 a kind of based on the FCM medical image cutting method for improving QPSO, it is characterised in that:Institute
The specific calculating step for stating step 3.3 is:Shown in the formula such as formula (5) to (9) that particle position updates, wherein pbiIt is i-th
The personal best particle of particle;Gb is the optimal location of entire population;Rand (0,1) is a function, and return value is one
Random number between [0,1];T is current evolutionary generation;tmaxIt is the maximum evolutionary generation of algorithm;β is known as shrinkage expansion system
Number, it is gradually reduced with the iteration of algorithm;piIt is the position of i-th of particle.New fit is calculated using the particle newly obtained
Angle value is answered, if reaching termination condition i.e. fitness value whether to be less than threshold value e or the number of iterations is more than M, otherwise continues to update
Particle.
A=rand (0,1) (5)
P=a*pbi+(1-a)*gb (6)
U=rand (0,1) (8)
。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810677834.4A CN108898602A (en) | 2018-06-27 | 2018-06-27 | A kind of FCM medical image cutting method based on improvement QPSO |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810677834.4A CN108898602A (en) | 2018-06-27 | 2018-06-27 | A kind of FCM medical image cutting method based on improvement QPSO |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108898602A true CN108898602A (en) | 2018-11-27 |
Family
ID=64346601
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810677834.4A Pending CN108898602A (en) | 2018-06-27 | 2018-06-27 | A kind of FCM medical image cutting method based on improvement QPSO |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108898602A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110111343A (en) * | 2019-05-07 | 2019-08-09 | 齐鲁工业大学 | A kind of middle intelligence image partition method and device based on improvement fuzzy C-mean algorithm |
CN110211103A (en) * | 2019-05-23 | 2019-09-06 | 电子科技大学 | Comentropy additivity based on infrared thermal imaging obscures defect characteristic and analyzes reconstructing method |
CN110751662A (en) * | 2019-10-17 | 2020-02-04 | 齐鲁工业大学 | Image segmentation method and system for quantum-behaved particle swarm optimization fuzzy C-means |
CN111189638A (en) * | 2019-12-24 | 2020-05-22 | 沈阳化工大学 | HMM and QPSO optimization algorithm-based bearing fault degree identification method |
CN111476303A (en) * | 2020-04-09 | 2020-07-31 | 国网河北省电力有限公司电力科学研究院 | Line loss analysis method of fuzzy C-means clustering based on quantum optimization particle swarm |
CN114841958A (en) * | 2022-05-04 | 2022-08-02 | 哈尔滨理工大学 | Method for automatically extracting image of brain subcutaneous tissue patch |
CN115295065A (en) * | 2022-10-09 | 2022-11-04 | 南京邮电大学 | Core grain test circuit based on flexible configurable module |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101923715A (en) * | 2010-09-02 | 2010-12-22 | 西安电子科技大学 | Image segmentation method based on texture information constrained clustering of particle swarm optimization space |
CN103824285A (en) * | 2014-01-27 | 2014-05-28 | 湖北工业大学 | Image segmentation method based on bat optimal fuzzy clustering |
CN105719293A (en) * | 2016-01-20 | 2016-06-29 | 东北大学 | Brain part MRI image segmentation method |
-
2018
- 2018-06-27 CN CN201810677834.4A patent/CN108898602A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101923715A (en) * | 2010-09-02 | 2010-12-22 | 西安电子科技大学 | Image segmentation method based on texture information constrained clustering of particle swarm optimization space |
CN103824285A (en) * | 2014-01-27 | 2014-05-28 | 湖北工业大学 | Image segmentation method based on bat optimal fuzzy clustering |
CN105719293A (en) * | 2016-01-20 | 2016-06-29 | 东北大学 | Brain part MRI image segmentation method |
Non-Patent Citations (1)
Title |
---|
武翠霞: "基于粒子群的图像分割算法及其应用", 《中国优秀硕士论文全文数据库》 * |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110111343A (en) * | 2019-05-07 | 2019-08-09 | 齐鲁工业大学 | A kind of middle intelligence image partition method and device based on improvement fuzzy C-mean algorithm |
CN110111343B (en) * | 2019-05-07 | 2021-08-31 | 齐鲁工业大学 | Middle-intelligence image segmentation method and device based on improved fuzzy C-means |
CN110211103A (en) * | 2019-05-23 | 2019-09-06 | 电子科技大学 | Comentropy additivity based on infrared thermal imaging obscures defect characteristic and analyzes reconstructing method |
CN110211103B (en) * | 2019-05-23 | 2022-03-25 | 电子科技大学 | Information entropy additive fuzzy defect characteristic analysis reconstruction method based on infrared thermal imaging |
CN110751662A (en) * | 2019-10-17 | 2020-02-04 | 齐鲁工业大学 | Image segmentation method and system for quantum-behaved particle swarm optimization fuzzy C-means |
CN110751662B (en) * | 2019-10-17 | 2022-10-25 | 齐鲁工业大学 | Image segmentation method and system for quantum-behaved particle swarm optimization fuzzy C-means |
CN111189638A (en) * | 2019-12-24 | 2020-05-22 | 沈阳化工大学 | HMM and QPSO optimization algorithm-based bearing fault degree identification method |
CN111189638B (en) * | 2019-12-24 | 2021-08-06 | 沈阳化工大学 | HMM and QPSO optimization algorithm-based bearing fault degree identification method |
CN111476303A (en) * | 2020-04-09 | 2020-07-31 | 国网河北省电力有限公司电力科学研究院 | Line loss analysis method of fuzzy C-means clustering based on quantum optimization particle swarm |
CN114841958A (en) * | 2022-05-04 | 2022-08-02 | 哈尔滨理工大学 | Method for automatically extracting image of brain subcutaneous tissue patch |
CN115295065A (en) * | 2022-10-09 | 2022-11-04 | 南京邮电大学 | Core grain test circuit based on flexible configurable module |
CN115295065B (en) * | 2022-10-09 | 2022-12-13 | 南京邮电大学 | Core grain test circuit based on flexible configurable module |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108898602A (en) | A kind of FCM medical image cutting method based on improvement QPSO | |
CN110120033A (en) | Based on improved U-Net neural network three-dimensional brain tumor image partition method | |
CN106846317B (en) | Medical image retrieval method based on feature extraction and similarity matching | |
CN108664976B (en) | Super-pixel-based fuzzy spectral clustering brain tumor image automatic segmentation method | |
CN109934235A (en) | A kind of unsupervised abdominal CT sequence image multiple organ automatic division method simultaneously | |
CN107680110B (en) | Inner ear three-dimensional level set segmentation method based on statistical shape model | |
CN106651874B (en) | Space domain splitting method after brain tumor surgery based on multi-modal MRI data | |
Taherdangkoo et al. | Segmentation of MR brain images using FCM improved by artificial bee colony (ABC) algorithm | |
CN113516653B (en) | Method for identifying glioma recurrence and necrosis through multi-feature fusion calculation | |
Bakas et al. | Segmentation of gliomas in multimodal magnetic resonance imaging volumes based on a hybrid generative-discriminative framework | |
Jiang et al. | An adaptive region growing based on neutrosophic set in ultrasound domain for image segmentation | |
Shelke et al. | Automated segmentation and detection of brain tumor from MRI | |
Tseng et al. | An adaptive thresholding method for automatic lung segmentation in CT images | |
CN104915989B (en) | Blood vessel three-dimensional dividing method based on CT images | |
Subashini et al. | Brain tumour detection using Pulse coupled neural network (PCNN) and back propagation network | |
Shia et al. | Intra-operative tumour margin evaluation in breast-conserving surgery with deep learning | |
Huang et al. | The segmentation of liver and vessels in CT images using 3D hierarchical seeded region growing | |
CN105469432B (en) | A kind of Automatic medical image segmentation method based on the tree-shaped department pattern of improvement | |
CN108346140A (en) | Based on the Otsu lung images dividing methods for improving PSO | |
Song et al. | Liver segmentation based on SKFCM and improved GrowCut for CT images | |
Li et al. | A hybrid approach to detection of brain hemorrhage candidates from clinical head ct scans | |
Dawod et al. | Adaptive Slices in Brain Haemorrhage Segmentation Based on the SLIC Algorithm. | |
Wang et al. | A study on the application of fuzzy information seeded region growing in brain MRI tissue segmentation | |
Menon et al. | CCS-GAN: COVID-19 CT-scan classification with very few positive training images | |
Liu et al. | Automatic liver segmentation using U-net in the assistance of CNN |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181127 |
|
RJ01 | Rejection of invention patent application after publication |