CN112419267A - Brain glioma segmentation model and method based on deep learning - Google Patents
Brain glioma segmentation model and method based on deep learning Download PDFInfo
- Publication number
- CN112419267A CN112419267A CN202011318476.1A CN202011318476A CN112419267A CN 112419267 A CN112419267 A CN 112419267A CN 202011318476 A CN202011318476 A CN 202011318476A CN 112419267 A CN112419267 A CN 112419267A
- Authority
- CN
- China
- Prior art keywords
- decoding module
- output end
- coding
- layer
- brain glioma
- 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
- 230000011218 segmentation Effects 0.000 title claims abstract description 58
- 201000007983 brain glioma Diseases 0.000 title claims abstract description 46
- 238000013135 deep learning Methods 0.000 title claims abstract description 21
- 238000000034 method Methods 0.000 title claims abstract description 11
- 230000009191 jumping Effects 0.000 claims abstract description 11
- 238000011176 pooling Methods 0.000 claims abstract description 6
- 238000012360 testing method Methods 0.000 claims description 21
- 238000012545 processing Methods 0.000 claims description 17
- 238000005481 NMR spectroscopy Methods 0.000 claims description 14
- 238000010606 normalization Methods 0.000 claims description 11
- 208000032612 Glial tumor Diseases 0.000 claims description 8
- 206010018338 Glioma Diseases 0.000 claims description 8
- 210000004556 brain Anatomy 0.000 claims description 6
- 239000000084 colloidal system Substances 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 6
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000012549 training Methods 0.000 claims description 3
- 230000003190 augmentative effect Effects 0.000 claims 1
- 238000003745 diagnosis Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 3
- 238000002591 computed tomography Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 208000003174 Brain Neoplasms Diseases 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000011157 brain segmentation Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- 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
- G06T2207/30096—Tumor; Lesion
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The invention discloses a brain glioma segmentation model and a segmentation method based on deep learning, belongs to the technical field of brain glioma segmentation, and aims to solve the technical problem of accurately segmenting brain glioma. A segmentation model comprising: the coding model comprises a convolutional layer and N coding modules, and comprises a hole dense unit and a pooling layer positioned at the output end of the hole dense unit; the decoding module comprises a head end decoding module, a middle decoding module and a tail end decoding module which are connected in sequence; the output end of the head end decoding module is in jumping connection with the output end of the hollow dense unit in the tail end coding module through a jumping connection layer; the output end of each intermediate decoding module is connected with the output end of the hollow dense unit in the corresponding intermediate coding module in a jumping way through a jumping connection layer; the tail end decoding module comprises a convolution layer, a hollow dense unit and a convolution layer which are connected in sequence, and the tail end decoding module is positioned at the output end of the jump layer related to the last decoding module.
Description
Technical Field
The invention relates to the technical field of glioma segmentation, in particular to a glioma segmentation model and a glioma segmentation method based on deep learning.
Background
In the current society, people are also paying more attention to their living conditions and medical conditions, and health is also a topic of great interest in people's life, so that medical science must be developed continuously, and medical images as important means for diagnosis and treatment of doctors must be developed correspondingly. Today, medical images play an important role in medical diagnosis that cannot be underestimated.
In medical images including nuclear Magnetic Resonance (MRI) and Computed Tomography (CT), medical image segmentation has a wide application in medical research, disease analysis, and surgical planning. The intelligent segmentation of brain glioma is of great significance to disease diagnosis and clinical decision.
The brain tumors are irregular in size and shape, uncertain in position and fuzzy in boundary, and the automatic segmentation of the brain glioma based on nuclear magnetic resonance is a challenging task. In clinic, the diagnosis of brain glioma is usually divided manually by experts, which is time-consuming and difficult, and the separation of brain glioma is the key of clinical treatment.
How to accurately segment the brain glioma is a technical problem to be solved.
Disclosure of Invention
The invention aims to overcome the defects and provides a brain glioma segmentation model and a segmentation method based on deep learning to solve the problem of how to accurately segment the brain glioma.
In a first aspect, the present invention provides a brain glioma segmentation model based on deep learning, including:
the coding model comprises a convolutional layer and N coding modules, wherein the convolutional layer is used for extracting features, the coding modules comprise a hole intensive unit and a pooling layer which are sequentially connected, the hole intensive unit is used for extracting the features by expanding a reception field and accelerating information flow, the N coding modules are sequentially connected and positioned at the output end of the convolutional layer, the coding module positioned at the transmission end is an end coding module, other coding modules are middle coding modules, and N is more than or equal to 3;
the decoding module comprises a head end decoding module, a middle decoding module and a tail end decoding module which are connected in sequence;
the head end decoding module comprises a hollow dense unit and an up-sampling convolution layer which are connected in sequence, the head end decoding module is positioned at the output end of the tail end coding module, and the output end of the head end decoding module is in jumping connection with the output end of the hollow dense unit in the tail end coding module through a jumping connection layer;
the number of the intermediate decoding modules is N-2, the intermediate decoding modules correspond to the intermediate coding modules one by one, each intermediate decoding module comprises a convolution layer, a space intensive unit and an up-sampling convolution layer which are sequentially connected, the output end of each intermediate decoding module is in jump connection with the output end of the hollow hole intensive unit in the intermediate coding module corresponding to the intermediate decoding module through a jump connection layer, and each intermediate decoding module is connected with the output end of the jump connection layer related to the previous decoding module;
the tail end decoding module comprises a convolution layer, a hollow dense unit and a convolution layer which are connected in sequence, and the tail end decoding module is positioned at the output end of the jump layer related to the last decoding module.
Preferably, the hole dense unit is composed of hole convolution layers and dense layer connections, and is used for extracting features by enlarging a receptive field and accelerating information flow.
Preferably, each of the void-dense cells is increased in expansion rate.
In a second aspect, the present invention provides a brain glioma segmentation method based on deep learning, including the following steps:
acquiring various three-dimensional nuclear magnetic resonance images as input images, wherein the nuclear magnetic resonance images comprise various MRI modalities and expert-labeled brain colloid regions;
normalizing the input image to make the input image conform to normal distribution;
constructing a brain glioma segmentation model based on deep learning according to any one of the first aspect;
training the brain glioma segmentation model based on the input image after normalization processing to obtain a trained brain glioma segmentation model;
acquiring a nuclear magnetic resonance image to be detected as a test image, wherein the nuclear magnetic resonance image comprises a plurality of MRI modalities and a brain colloid region marked by an expert;
carrying out normalization processing on the test image to enable the test image to conform to normal distribution;
and (4) carrying out segmentation processing on the test image through the trained brain glioma segmentation model.
Preferably, the input image after the normalization processing is subjected to size cutting, and the brain glioma segmentation model is trained on the basis of the input image after the size cutting, so that a trained brain glioma segmentation model is obtained;
and cutting the size of the test image subjected to the normalization processing, and carrying out segmentation processing on the test image subjected to the size cutting through the trained glioma segmentation model based on the test image subjected to the size cutting.
The brain glioma segmentation model and the segmentation method based on deep learning have the following advantages:
1. the coding module and the decoding module both comprise a hole dense unit, and the model can extract the characteristics of the multi-scale receptive field through the hole dense unit, improve the characteristic extraction capability and improve the segmentation effect;
2. the hole convolution expands the receiving domain to capture context information, and the dense connection maximizes the gradient information flow, so that the network can better learn the representation of the image;
3. the hole-dense cells are connected to the corresponding decoding modules using jumping connection layers, which improve the information flow in the model.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed for the embodiments or the prior art descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.
The invention is further described below with reference to the accompanying drawings.
Fig. 1 is a schematic structural diagram of a brain glioma segmentation model based on deep learning in embodiment 1;
fig. 2 is a schematic structural diagram of a hole dense unit in a brain glioma segmentation model based on deep learning in embodiment 1.
Detailed Description
The present invention is further described in the following with reference to the drawings and the specific embodiments so that those skilled in the art can better understand the present invention and can implement the present invention, but the embodiments are not to be construed as limiting the present invention, and the embodiments and the technical features of the embodiments can be combined with each other without conflict.
The embodiment of the invention provides a brain glioma segmentation model and a segmentation method based on deep learning, which are used for solving the technical problem of accurately segmenting brain glioma.
Example 1:
the invention relates to a brain glioma segmentation model based on deep learning, which comprises a coding model and a decoding model, wherein the coding model comprises a convolution layer and N coding modules, the coding comprises a hole dense unit and a pooling layer positioned at the output end of the hole dense unit, the N coding modules are sequentially connected and positioned at the output end of the convolution layer, the coding module positioned at the transmission tail end is a tail end coding module, other coding modules are middle coding modules, and N is more than or equal to 3; the decoding module comprises a head end decoding module, a middle decoding module and a tail end decoding module which are connected in sequence; the head end decoding module comprises a hollow dense unit and an up-sampling convolutional layer which are sequentially connected, the head end decoding module is positioned at the output end of the tail end coding module, and the output end of the head end decoding module is in jumping connection with the output end of the hollow dense unit in the tail end coding module through a jumping connection layer; the number of the middle decoding modules is N-2, the middle decoding modules correspond to the middle coding modules one by one, each middle decoding module comprises a convolution layer, a space intensive unit and an up-sampling convolution layer which are sequentially connected, the output end of each middle decoding module is in jump connection with the output end of the hollow hole intensive unit in the middle coding module corresponding to the middle decoding module through a jump connection layer, and each middle decoding module is connected with the output end of the jump connection layer related to the previous decoding module; the tail end decoding module comprises a convolution layer, a hollow dense unit and a convolution layer which are connected in sequence, and the tail end decoding module is positioned at the output end of the jump layer related to the last decoding module.
In this embodiment, there are three coding modules, where the convolution layer is a 3 × 3 convolution layer, and is used to extract effective features in the input image; the hole dense unit is added in the network to better promote gradient flow and enable the network to learn better feature representation of the image while expanding the receptive field; the pooling layer is the largest pooling layer of 2 x 2, reducing the image to half of the original image.
The number of the decoding modules is four, wherein the upsampling convolutional layer is 2 x 2, and the image is gradually restored to the size of the original image; the hole dense unit enables the decoding module to extract the characteristics of the multi-scale receptive field, and can improve the capability of characteristic extraction and the segmentation effect.
In this embodiment, the hole dense units are formed by hole convolution and dense connection, the hole dense units extract features by expanding a receptive field and accelerating information flow, the dense connection further improves the information flow between layers, the receptive field is expanded without reducing image resolution and introducing additional parameters and calculation methods, and an expansion rate is added to each hole dense unit in this embodiment; the kernel size of each convolution block is 3 × 3 × 3, and the expansion ratio of each hole-dense cell is {1,2,4,8 }. The hole convolution expands the acceptance domain to capture context information, and dense connections maximize the flow of gradient information, enabling the network to better learn the representation of the image.
In the brain glioma segmentation model of the embodiment, the hole dense units of each coding module are connected to the corresponding decoding module by using jump connection layers, and the jump connections promote gradient flow in the model and information transmission between the coding module and the decoding module. The model can extract multi-scale features and fuse high-level semantic information, so that the network has better feature extraction capability on the 3D glioma image.
Example 2:
the brain glioma segmentation method based on deep learning comprises the following steps:
s100, acquiring various three-dimensional nuclear magnetic resonance images as input images, wherein the nuclear magnetic resonance images comprise various MRI modalities and expert-labeled brain colloid regions;
s200, normalizing the input image to enable the input image to be in accordance with normal distribution;
s300, constructing a brain glioma segmentation model based on deep learning disclosed in the embodiment 1;
s400, training the brain glioma segmentation model based on the input image after normalization processing to obtain a trained brain glioma segmentation model;
s500, acquiring a nuclear magnetic resonance image to be detected as a test image, wherein the nuclear magnetic resonance image comprises a plurality of MRI modalities and a brain colloid area marked by an expert;
s600, normalizing the test image to enable the test image to be in accordance with normal distribution;
and S700, carrying out segmentation processing on the test image through a trained brain glioma segmentation model.
In this embodiment, the experimental data adopts a public dataset brats (multimodal brain segmentation change) provided by MICCAI and containing expert-labeled brain glioma, and the images of each patient include 4 MRI modalities and expert-labeled brain glioma regions, and the images are normalized so that the input images conform to normal distribution, that is, the average value of the images is 0 and the standard deviation is 1. The normalized image has stronger stability and can accelerate network convergence.
As an improvement, because an original image is large and hardware resources such as a video memory and a memory are limited, the input image after normalization processing is subjected to size cutting to be cut into 128 × 128 × 128, and the brain glioma segmentation model is trained on the basis of the input image after size cutting to obtain a trained brain glioma segmentation model; correspondingly, the test image after the normalization processing is subjected to size cutting, the test image is cut into 128 x 128, and based on the test image after the size cutting, the segmentation processing is carried out through a trained glioma segmentation model.
While the invention has been shown and described in detail in the drawings and in the preferred embodiments, it is not intended to limit the invention to the embodiments disclosed, and it will be apparent to those skilled in the art that many more embodiments of the invention are possible that combine the features of the different embodiments described above and still fall within the scope of the invention.
Claims (5)
1. Brain glioma segmentation model based on deep learning is characterized by comprising:
the coding model comprises a convolutional layer and N coding modules, wherein the convolutional layer is used for extracting features, the codes comprise a hole dense unit and a pooling layer located at the output end of the hole dense unit, the hole dense unit is used for extracting the features by enlarging the receptive field and accelerating the information flow, the N coding modules are sequentially connected and located at the output end of the convolutional layer, the coding module located at the transmission tail end is a tail end coding module, the other coding modules are middle coding modules, and N is more than or equal to 3;
the decoding module comprises a head end decoding module, a middle decoding module and a tail end decoding module which are connected in sequence;
the head end decoding module comprises a hollow dense unit and an up-sampling convolution layer which are connected in sequence, the head end decoding module is positioned at the output end of the tail end coding module, and the output end of the head end decoding module is in jumping connection with the output end of the hollow dense unit in the tail end coding module through a jumping connection layer;
the number of the intermediate decoding modules is N-2, the intermediate decoding modules correspond to the intermediate coding modules one by one, each intermediate decoding module comprises a convolution layer, a space intensive unit and an up-sampling convolution layer which are sequentially connected, the output end of each intermediate decoding module is in jump connection with the output end of the hollow hole intensive unit in the intermediate coding module corresponding to the intermediate decoding module through a jump connection layer, and each intermediate decoding module is connected with the output end of the jump connection layer related to the previous decoding module;
the tail end decoding module comprises a convolution layer, a hollow dense unit and a convolution layer which are connected in sequence, and the tail end decoding module is positioned at the output end of the jump layer related to the last decoding module.
2. The deep learning based glioma segmentation model of claim 1 wherein the hole dense cells are composed of hole convolution layers and dense layer connections for extracting features by enlarging the receptive field and speeding up the information flow.
3. The deep learning based brain glioma segmentation model of claim 2 wherein each hole dense cell is augmented with an expansion rate.
4. The brain glioma segmentation method based on deep learning is characterized by comprising the following steps:
acquiring various three-dimensional nuclear magnetic resonance images as input images, wherein the nuclear magnetic resonance images comprise various MRI modalities and expert-labeled brain colloid regions;
normalizing the input image to make the input image conform to normal distribution;
constructing a brain glioma segmentation model based on deep learning according to any one of claims 1 to 3;
training the brain glioma segmentation model based on the input image after normalization processing to obtain a trained brain glioma segmentation model;
acquiring a nuclear magnetic resonance image to be detected as a test image, wherein the nuclear magnetic resonance image comprises a plurality of MRI modalities and a brain colloid region marked by an expert;
carrying out normalization processing on the test image to enable the test image to conform to normal distribution;
and (4) carrying out segmentation processing on the test image through the trained brain glioma segmentation model.
5. The brain glioma segmentation method based on deep learning of claim 4, wherein the input image after the normalization processing is subjected to size clipping, and the brain glioma segmentation model is trained based on the input image after the size clipping to obtain a trained brain glioma segmentation model;
and cutting the size of the test image subjected to the normalization processing, and carrying out segmentation processing on the test image subjected to the size cutting through the trained glioma segmentation model based on the test image subjected to the size cutting.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011318476.1A CN112419267A (en) | 2020-11-23 | 2020-11-23 | Brain glioma segmentation model and method based on deep learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011318476.1A CN112419267A (en) | 2020-11-23 | 2020-11-23 | Brain glioma segmentation model and method based on deep learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112419267A true CN112419267A (en) | 2021-02-26 |
Family
ID=74777317
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011318476.1A Pending CN112419267A (en) | 2020-11-23 | 2020-11-23 | Brain glioma segmentation model and method based on deep learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112419267A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112991355A (en) * | 2021-05-13 | 2021-06-18 | 南京应用数学中心 | 3D brain lesion segmentation method based on optimal transmission |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2019200747A1 (en) * | 2018-04-20 | 2019-10-24 | 平安科技(深圳)有限公司 | Method and device for segmenting proximal femur, computer apparatus, and storage medium |
CN110619641A (en) * | 2019-09-02 | 2019-12-27 | 南京信息工程大学 | Automatic segmentation method of three-dimensional breast cancer nuclear magnetic resonance image tumor region based on deep learning |
CN110866931A (en) * | 2019-11-18 | 2020-03-06 | 东声(苏州)智能科技有限公司 | Image segmentation model training method and classification-based enhanced image segmentation method |
CN111145170A (en) * | 2019-12-31 | 2020-05-12 | 电子科技大学 | Medical image segmentation method based on deep learning |
CN111354002A (en) * | 2020-02-07 | 2020-06-30 | 天津大学 | Kidney and kidney tumor segmentation method based on deep neural network |
CN111369563A (en) * | 2020-02-21 | 2020-07-03 | 华南理工大学 | Semantic segmentation method based on pyramid void convolutional network |
WO2020215236A1 (en) * | 2019-04-24 | 2020-10-29 | 哈尔滨工业大学(深圳) | Image semantic segmentation method and system |
CN111862049A (en) * | 2020-07-22 | 2020-10-30 | 齐鲁工业大学 | Brain glioma segmentation network system and segmentation method based on deep learning |
-
2020
- 2020-11-23 CN CN202011318476.1A patent/CN112419267A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2019200747A1 (en) * | 2018-04-20 | 2019-10-24 | 平安科技(深圳)有限公司 | Method and device for segmenting proximal femur, computer apparatus, and storage medium |
WO2020215236A1 (en) * | 2019-04-24 | 2020-10-29 | 哈尔滨工业大学(深圳) | Image semantic segmentation method and system |
CN110619641A (en) * | 2019-09-02 | 2019-12-27 | 南京信息工程大学 | Automatic segmentation method of three-dimensional breast cancer nuclear magnetic resonance image tumor region based on deep learning |
CN110866931A (en) * | 2019-11-18 | 2020-03-06 | 东声(苏州)智能科技有限公司 | Image segmentation model training method and classification-based enhanced image segmentation method |
CN111145170A (en) * | 2019-12-31 | 2020-05-12 | 电子科技大学 | Medical image segmentation method based on deep learning |
CN111354002A (en) * | 2020-02-07 | 2020-06-30 | 天津大学 | Kidney and kidney tumor segmentation method based on deep neural network |
CN111369563A (en) * | 2020-02-21 | 2020-07-03 | 华南理工大学 | Semantic segmentation method based on pyramid void convolutional network |
CN111862049A (en) * | 2020-07-22 | 2020-10-30 | 齐鲁工业大学 | Brain glioma segmentation network system and segmentation method based on deep learning |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112991355A (en) * | 2021-05-13 | 2021-06-18 | 南京应用数学中心 | 3D brain lesion segmentation method based on optimal transmission |
CN112991355B (en) * | 2021-05-13 | 2021-08-31 | 南京应用数学中心 | 3D brain lesion segmentation method based on optimal transmission |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110070540B (en) | Image generation method and device, computer equipment and storage medium | |
CN110232383A (en) | A kind of lesion image recognition methods and lesion image identifying system based on deep learning model | |
CN111932529B (en) | Image classification and segmentation method, device and system | |
CN112446891B (en) | Medical image segmentation method based on U-Net network brain glioma | |
CN110444277B (en) | Multi-mode brain MRI image bidirectional conversion method based on multi-generation and multi-confrontation | |
CN111986101B (en) | Cerebrovascular map construction method | |
CN112037171B (en) | Multi-mode feature fusion-based multi-task MRI brain tumor image segmentation method | |
CN114926477A (en) | Brain tumor multi-modal MRI (magnetic resonance imaging) image segmentation method based on deep learning | |
CN114782350A (en) | Multi-modal feature fusion MRI brain tumor image segmentation method based on attention mechanism | |
CN112529909A (en) | Tumor image brain region segmentation method and system based on image completion | |
CN111754532A (en) | Image segmentation model searching method and device, computer equipment and storage medium | |
CN112884788B (en) | Cup optic disk segmentation method and imaging method based on rich context network | |
CN113052856A (en) | Hippocampus three-dimensional semantic network segmentation method based on multi-scale feature multi-path attention fusion mechanism | |
CN113674291B (en) | Full-type aortic dissection true and false cavity image segmentation method and system | |
CN112288749A (en) | Skull image segmentation method based on depth iterative fusion depth learning model | |
CN114882048A (en) | Image segmentation method and system based on wavelet scattering learning network | |
CN116309615A (en) | Multi-mode MRI brain tumor image segmentation method | |
CN112419267A (en) | Brain glioma segmentation model and method based on deep learning | |
CN114119788A (en) | Multi-modal medical image coding and generating method based on countermeasure generation network | |
CN117058376A (en) | Brain image data processing method based on improved Unet++ | |
CN116958705A (en) | Medical image classifying system based on graph neural network | |
CN114463320B (en) | Magnetic resonance imaging brain glioma IDH gene prediction method and system | |
CN116739988A (en) | Deep learning cerebral hemorrhage classification method based on multi-difficulty course learning | |
CN116030043A (en) | Multi-mode medical image segmentation method | |
CN115841457A (en) | Three-dimensional medical image segmentation method fusing multi-view information |
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: 20210226 |
|
RJ01 | Rejection of invention patent application after publication |