CN113066081A - Breast tumor molecular subtype detection method based on three-dimensional MRI (magnetic resonance imaging) image - Google Patents
Breast tumor molecular subtype detection method based on three-dimensional MRI (magnetic resonance imaging) image Download PDFInfo
- Publication number
- CN113066081A CN113066081A CN202110422631.2A CN202110422631A CN113066081A CN 113066081 A CN113066081 A CN 113066081A CN 202110422631 A CN202110422631 A CN 202110422631A CN 113066081 A CN113066081 A CN 113066081A
- Authority
- CN
- China
- Prior art keywords
- tumor
- breast
- image
- molecular subtype
- model
- 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
- 208000026310 Breast neoplasm Diseases 0.000 title claims abstract description 36
- 206010006187 Breast cancer Diseases 0.000 title claims abstract description 34
- 238000001514 detection method Methods 0.000 title claims abstract description 15
- 238000002595 magnetic resonance imaging Methods 0.000 title description 24
- 206010028980 Neoplasm Diseases 0.000 claims abstract description 41
- 238000012549 training Methods 0.000 claims abstract description 27
- 210000000481 breast Anatomy 0.000 claims abstract description 14
- 238000000034 method Methods 0.000 claims abstract description 14
- 238000012545 processing Methods 0.000 claims abstract description 8
- 239000000284 extract Substances 0.000 claims abstract description 4
- 238000012360 testing method Methods 0.000 claims description 16
- 238000013527 convolutional neural network Methods 0.000 claims description 13
- 230000011218 segmentation Effects 0.000 claims description 11
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 6
- 230000009466 transformation Effects 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 5
- 238000013507 mapping Methods 0.000 claims description 5
- 101001012157 Homo sapiens Receptor tyrosine-protein kinase erbB-2 Proteins 0.000 claims description 4
- 102100030086 Receptor tyrosine-protein kinase erbB-2 Human genes 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 238000009499 grossing Methods 0.000 claims description 3
- 208000026535 luminal A breast carcinoma Diseases 0.000 claims description 3
- 208000026534 luminal B breast carcinoma Diseases 0.000 claims description 3
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 2
- 238000011176 pooling Methods 0.000 claims description 2
- 238000004393 prognosis Methods 0.000 abstract description 4
- 238000012216 screening Methods 0.000 abstract description 2
- 238000011282 treatment Methods 0.000 abstract description 2
- 238000013528 artificial neural network Methods 0.000 abstract 1
- 238000003745 diagnosis Methods 0.000 abstract 1
- 239000003814 drug Substances 0.000 abstract 1
- 229940079593 drug Drugs 0.000 abstract 1
- 230000000472 traumatic effect Effects 0.000 abstract 1
- 230000006870 function Effects 0.000 description 11
- 230000014509 gene expression Effects 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000011337 individualized treatment Methods 0.000 description 2
- 238000012937 correction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012252 genetic analysis Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 108090000623 proteins and genes Proteins 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000009885 systemic effect Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 229940126585 therapeutic drug Drugs 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20092—Interactive image processing based on input by user
- G06T2207/20104—Interactive definition of region of interest [ROI]
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
- Image Analysis (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Image Processing (AREA)
Abstract
A breast tumor molecular subtype detection method based on three-dimensional MRI images belongs to the technical field of computer image processing; the method specifically comprises the steps of training a full convolution FCN model, utilizing a 3D U-Net framework to segment and extract tumor regions in a breast MRI image, taking the segmented tumor region image as input, transmitting the segmented tumor region image into a proposed 3D convolution neural network, and outputting the probability that the tumor belongs to four types through a softmax function, so that the molecular subtype of the tumor is predicted. The method automatically extracts the tumor regions in the image, and utilizes the 3D convolution network to evaluate the molecular subtype of each tumor region. Histopathological examination is the "gold standard" for breast cancer diagnosis, but is very traumatic to the patient, and requires a certain time for detection, which affects the prognosis of the patient. The breast tumor molecular subtype detection based on the three-dimensional MRI image discovers and identifies the molecular subtype of the breast tumor in the primary screening stage, can assist doctors to select targeted treatment drugs, and has important significance for improving the prognosis of patients.
Description
Technical Field
The invention belongs to the technical field of computer image processing, and relates to a breast tumor molecular subtype detection method based on a three-dimensional MRI image.
Background
With the development of molecular biology, gene expression profiles and gene chip technology are widely used in the research of breast cancer, which is considered as a single disease with multiple subtypes. The natural course of disease and the response to systemic or local treatment vary from one breast cancer to another, and these high degrees of heterogeneity suggest the need for individualized treatment of breast cancers of different molecular classifications. The breast tumor is divided into four molecular subtypes of Luminal-A type, Luminal-B type, HER2+ type and Basal-like type according to a gene expression profile, a useful basis is provided for clinical selection of individualized treatment, and the gene expression profile has important significance for prognosis of early breast cancer.
However, the determination of breast cancer subtypes by genetic analysis is invasive and costly, requiring specialized equipment and technical expertise. Recent research and clinical research prove that the correlation between the breast cancer subtype and the deep learning characteristics is realized, the deep learning characteristics are extracted from an MRI image in the primary screening stage of a patient to detect the molecular subtype of the breast cancer, a physician can be assisted to select a targeted therapeutic drug, and the method has important significance for improving the prognosis of the patient.
Disclosure of Invention
In view of the above-mentioned background, the present invention provides a breast tumor molecular subtype detection method based on three-dimensional MRI images, which is characterized by comprising the following steps:
a, preprocessing a breast MRI image;
b, segmenting and extracting a tumor region in the breast MRI image by using an FCN model;
and C, predicting the molecular subtype of the tumor region image by using a CNN model.
2. The breast tumor molecular subtype detection method based on the three-dimensional MRI image according to claim 1, characterized in that the preprocessing of the breast MRI image in the step A includes removing artifacts and noise of the image, aligning the images at different time points, and temporally correcting and smoothing the signals acquired at different layers to obtain a higher signal-to-noise ratio;
3. the method for detecting molecular subtype of breast tumor based on three-dimensional MRI image according to claim 1, characterized in that the tumor region segmentation and extraction in the step B comprises the following steps:
a, selecting a convolutional neural network model structure, and acquiring shallow information such as physical outline, edge, texture and the like of a breast tumor and deep abstract information in an input image by using a 3D U-Net architecture to extract an interested region, namely a mask of the tumor;
b, respectively taking out breast areas with complete tumor marks and clear images from the MRI images in a sliding block cubic block mode, training and testing the tumor segmentation models, turning over the training images, mirroring the training images, performing affine transformation and other data enhancement processing;
and C, training the FCN, adjusting network parameters, selecting an optimizer (Adam), selecting a Dice coefficient difference function as a loss function, and optimizing a network model by continuously reducing the Dice coefficient difference function. The Dice coefficient is a metric function that calculates the similarity in two samples, and is:
wherein, YtrueTrue label value, Y, representing training datapredictRepresenting the predicted value of the current model in the forward iteration process. The Dice coefficient difference function is calculated as:
Loss＝1-S
and D, testing the FCN model by using the test set, and reserving the optimal segmentation model.
4. The method for detecting molecular subtype of breast tumor based on three-dimensional MRI image according to claim 1, characterized in that the tumor molecular subtype classification extraction in the step C comprises the following steps:
a, selecting a model structure of a convolutional neural network, acquiring image information of a tumor region by using a 3D resnet18 framework, and performing molecular subtype classification on a breast tumor;
b, dividing the image into four types of a luminal-A type, a luminal-B type, a HER2+ type and a Basal-like type according to molecular subtypes for training and testing tumor typing, rotating and mirroring the training set image, and performing data enhancement processing such as affine transformation and the like;
and C, training the 3D resnet18 network and adjusting network parameters. The model comprises a network structure and an activation function, wherein the activation function is probability output of four types of tumors by using a softmax function at a network output layer. The softmax activation function is:
wherein z isiFor the output value of the ith node, C is the number of output nodes, namely the number of classification nodes, and the output value of the multi-classification can be converted into the value of [0, 1] through the softmax function]Interval and sum is 1. ReLU function is used between convolutional layer and pooling layer, which is:
f(x)＝max(0，x)
and D, testing the CNN model by using the test set, and reserving the optimal typing model.
5. The breast tumor molecular subtype detection method based on three-dimensional MRI image as claimed in claim 3, wherein in step B, the tumor segmentation data set extracts the breast region with 128 x 128 pixels by means of sliding block cubic block extraction, and converts the corresponding mark into a binary image, the tumor region in the mark line is represented by white, and the non-tumor region is represented by black.
6. The method for detecting molecular subtypes of breast tumors based on three-dimensional MRI images as claimed in claim 4, wherein in step C, block structures in the 3D resnet18 network are all replaced by topological block structures.
7. The breast tumor molecular subtype detection method based on the three-dimensional MRI image according to claim 6, characterized in that the topological block structure is formed by taking a plurality of groups of convolutions of each structural unit, respectively mapping and then summing the convolutions, and summing the mapped groups with the input. The mathematical formula is as follows:
where x is the input of the building block, y is the output of the building block, C is the number of convolution groups, Ti(x) A mapping for each convolution group.
Drawings
FIG. 1 is a flow chart of molecular subtype detection of breast tumor in accordance with the present invention
FIG. 2 is a flow chart of FCN network training of the present invention
FIG. 3 is a CNN network training flowchart of the present invention
FIG. 4 shows an original block structure and a topological block structure used in the present invention
Detailed Description
In order to make the technical problems and advantages of the present invention to be clearly understood, the following detailed description of the embodiments of the present invention is made with reference to the accompanying drawings, but it should be understood that the scope of the present invention is not limited by the embodiments.
A breast tumor molecular subtype detection method based on a three-dimensional MRI image is shown in figure 1, and comprises the following specific steps:
step S1: the breast MRI image to be detected is first pre-processed. Removing artifacts and noise of the images, aligning the images at different time points, and performing temporal correction and smoothing on signals acquired at different layers to obtain a higher signal-to-noise ratio;
step S2: and (3) carrying out breast MRI image segmentation by using an FCN model to extract a tumor region. The process of training the FCN model is shown in FIG. 2 and mainly comprises the following steps:
1. and respectively taking out the breast areas with complete tumor markers and clear images from the MRI images in a way of taking the cube by a sliding block according to the ratio of 8: 2, the proportion is used for training and testing the tumor segmentation model, and data enhancement processing such as turning, mirroring and affine transformation is carried out on a training image;
2. inputting the training set image after data enhancement into a 3D U-Net network to train the network, and keeping a model which is best represented on a segmentation test set;
3. taking the image to be detected preprocessed in the step S1 as input, and transmitting the input image into an optimal FCN model to perform segmentation operation to obtain a tumor area image of the image to be detected;
step S3: the molecular typing work of the tumor region MRI image is carried out by using the improved CNN model, and the process of training the CNN model is shown in figure 3 and mainly comprises the following steps:
1. the images are divided into four types of molecular-A type, molecular-B type, HER2+ type and Basal-like type according to molecular subtypes for training and testing tumor typing. Rotating and mirroring the training set image, and performing affine transformation and other data enhancement processing;
2. inputting the preprocessed MRI images of the training set into an improved 3D Resnet18 network for training, and keeping a model which best represents on the test set; the improvement scheme of the fractal network is that the original block structure of the 3D Resnet18 is replaced by a topological block structure, and the traditional block structure and the topological block structure are shown in FIG. 4. The topological block structure is characterized in that each structural unit takes a plurality of groups of convolutions, the convolutions are respectively mapped and summed, and then the summed value is summed with input. The mathematical formula is as follows:
where x is the input of the building block, y is the output of the building block, C is the number of convolution groups, Ti(x) A mapping for each convolution group.
3. And (5) taking the tumor region of the image to be detected generated in the step (S2) as an input, transmitting the tumor region into the optimal CNN model to perform subtype typing operation, and outputting the breast tumor molecular subtype of the MRI image to be detected.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements, etc. made within the spirit and principles of the invention shall be included in the scope of the present invention.
Claims (7)
1. The breast tumor molecular subtype detection method based on the three-dimensional MRI image is characterized by comprising the following steps of:
a, preprocessing a breast MRI image to be detected;
b, segmenting and extracting a tumor region in the breast MRI image by using an FCN model;
and C, predicting the molecular subtype of the tumor region image by using a CNN model.
2. The method for detecting molecular subtype of breast tumor based on three-dimensional MRI image according to claim 1, characterized in that the preprocessing of the breast MRI image in the step a includes removing artifacts and noise of the image, aligning the images at different time points, and temporally correcting and smoothing the signals acquired at different slices to obtain higher signal-to-noise ratio.
3. The method for detecting molecular subtype of breast tumor based on three-dimensional MRI image according to claim 1, characterized in that the tumor region segmentation and extraction in the step B comprises the following steps:
a, selecting a convolutional neural network model structure, and acquiring shallow information such as physical outline, edge, texture and the like of a breast tumor and deep abstract information in an input image by using a 3D U-Net architecture to extract an interested region, namely a mask of the tumor;
b, taking out a breast area with complete tumor markers and clear images from the MRI image in a sliding block cubic block mode, training and testing a tumor segmentation model, rotating and mirroring the training set image, and performing affine transformation and other data enhancement processing;
and C, training the FCN, adjusting network parameters, selecting an optimizer (Adam), selecting a Dice coefficient difference function as a loss function, and optimizing a network model by continuously reducing the Dice coefficient difference function. The Dice coefficient is a metric function that calculates the similarity in two samples, and is:
wherein, YtureTrue label value, Y, representing training datapredictRepresenting the predicted value of the current model in the forward iteration process. The Dice coefficient difference function is calculated as:
Loss＝1-S
and D, testing the FCN model by using the test set, and reserving the optimal segmentation model.
4. The method for detecting molecular subtype of breast tumor based on three-dimensional MRI image according to claim 1, characterized in that said molecular subtype of tumor in step C comprises the following steps:
a, selecting a model structure of a convolutional neural network, acquiring image information of a tumor region by using a 3D resnet18 framework, and performing molecular subtype classification on a breast tumor;
b, dividing the images into four types of Luminal-A type, Luminal-B type, HER2+ type and Basal-like type according to molecular subtypes for training and testing tumor typing, rotating and mirroring the images of the training set, and performing data enhancement processing such as affine transformation and the like;
and C, training the 3D resnet18 network and adjusting network parameters. The CNN model comprises a network structure and an activation function, wherein the activation function is the probability output of four types of tumors by using a softmax function at a network output layer. The softmax activation function is:
wherein zi is the output value of the ith node, C is the number of output nodes, i.e. the number of classification nodes, and the output values of multiple classifications can be converted into probability distribution with a sum of 1 in the interval of [0, 1] by a softmax function. ReLU function is used between convolutional layer and pooling layer, which is:
f(x)＝max(0，x)
and D, testing the CNN model by using the test set, and reserving the optimal typing model.
5. The breast tumor molecular subtype detection method based on three-dimensional MRI image as claimed in claim 3, wherein in step B, the tumor segmentation data set extracts the breast region with 128 x 128 pixels by means of sliding block cubic block extraction, and converts the corresponding mark into a binary image, the tumor region in the mark line is represented by white, and the non-tumor region is represented by black.
6. The method for detecting molecular subtypes of breast tumors based on three-dimensional MRI images as claimed in claim 4, wherein in step C, block structures in the 3D resnet18 network are all replaced by topological block structures.
7. The breast tumor molecular subtype detection method based on the three-dimensional MRI image according to claim 6, characterized in that the topological block structure is formed by taking a plurality of groups of convolutions of each structural unit, respectively mapping and then summing the convolutions, and summing the mapped groups with the input. The mathematical formula is as follows:
where x is the input of the building block, y is the output of the building block, C is the number of convolution groups, Ti(x) A mapping for each convolution group.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110422631.2A CN113066081B (en) | 2021-04-15 | 2021-04-15 | Mammary tumor molecular subtype detection method based on three-dimensional MRI image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110422631.2A CN113066081B (en) | 2021-04-15 | 2021-04-15 | Mammary tumor molecular subtype detection method based on three-dimensional MRI image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113066081A true CN113066081A (en) | 2021-07-02 |
| CN113066081B CN113066081B (en) | 2023-07-18 |
Family
ID=76567017
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110422631.2A Active CN113066081B (en) | 2021-04-15 | 2021-04-15 | Mammary tumor molecular subtype detection method based on three-dimensional MRI image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113066081B (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107464250A (en) * | 2017-07-03 | 2017-12-12 | 深圳市第二人民医院 | Tumor of breast automatic division method based on three-dimensional MRI image |
| CN109215035A (en) * | 2018-07-16 | 2019-01-15 | 江南大学 | A kind of brain MRI hippocampus three-dimensional dividing method based on deep learning |
| CN109859190A (en) * | 2019-01-31 | 2019-06-07 | 北京工业大学 | A kind of target area detection method based on deep learning |
| CN109978838A (en) * | 2019-03-08 | 2019-07-05 | 腾讯科技（深圳）有限公司 | Image-region localization method, device and Medical Image Processing equipment |
| CN110265141A (en) * | 2019-05-13 | 2019-09-20 | 上海大学 | A kind of liver neoplasm CT images computer aided diagnosing method |
| CN110942464A (en) * | 2019-11-08 | 2020-03-31 | 浙江工业大学 | PET image segmentation method fusing 2-dimensional and 3-dimensional models |
| CN111402226A (en) * | 2020-03-13 | 2020-07-10 | 浙江工业大学 | Surface defect detection method based on cascade convolution neural network |
| CN111798469A (en) * | 2020-07-13 | 2020-10-20 | 珠海函谷科技有限公司 | Digital image small data set semantic segmentation method based on deep convolutional neural network |
| CN112070685A (en) * | 2020-08-10 | 2020-12-11 | 武汉大学 | Method for predicting dynamic soft tissue motion of HIFU treatment system |
| CN112241679A (en) * | 2020-09-14 | 2021-01-19 | 浙江理工大学 | Automatic garbage classification method |
| CN112348769A (en) * | 2020-08-20 | 2021-02-09 | 盐城工学院 | Intelligent kidney tumor segmentation method and device in CT (computed tomography) image based on U-Net depth network model |
-
2021
- 2021-04-15 CN CN202110422631.2A patent/CN113066081B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107464250A (en) * | 2017-07-03 | 2017-12-12 | 深圳市第二人民医院 | Tumor of breast automatic division method based on three-dimensional MRI image |
| CN109215035A (en) * | 2018-07-16 | 2019-01-15 | 江南大学 | A kind of brain MRI hippocampus three-dimensional dividing method based on deep learning |
| CN109859190A (en) * | 2019-01-31 | 2019-06-07 | 北京工业大学 | A kind of target area detection method based on deep learning |
| CN109978838A (en) * | 2019-03-08 | 2019-07-05 | 腾讯科技（深圳）有限公司 | Image-region localization method, device and Medical Image Processing equipment |
| CN110265141A (en) * | 2019-05-13 | 2019-09-20 | 上海大学 | A kind of liver neoplasm CT images computer aided diagnosing method |
| CN110942464A (en) * | 2019-11-08 | 2020-03-31 | 浙江工业大学 | PET image segmentation method fusing 2-dimensional and 3-dimensional models |
| CN111402226A (en) * | 2020-03-13 | 2020-07-10 | 浙江工业大学 | Surface defect detection method based on cascade convolution neural network |
| CN111798469A (en) * | 2020-07-13 | 2020-10-20 | 珠海函谷科技有限公司 | Digital image small data set semantic segmentation method based on deep convolutional neural network |
| CN112070685A (en) * | 2020-08-10 | 2020-12-11 | 武汉大学 | Method for predicting dynamic soft tissue motion of HIFU treatment system |
| CN112348769A (en) * | 2020-08-20 | 2021-02-09 | 盐城工学院 | Intelligent kidney tumor segmentation method and device in CT (computed tomography) image based on U-Net depth network model |
| CN112241679A (en) * | 2020-09-14 | 2021-01-19 | 浙江理工大学 | Automatic garbage classification method |
Non-Patent Citations (2)
| Title |
|---|
| RICHARD HA ET AL: "Predicting Breast Cancer Molecular Subtype with MRI Dataset Utilizing Convolutional Neural Network Algorithm", 《JOURNAL OF DIGITAL IMAGING》 * |
| RICHARD HA ET AL: "Predicting Breast Cancer Molecular Subtype with MRI Dataset Utilizing Convolutional Neural Network Algorithm", 《JOURNAL OF DIGITAL IMAGING》, 31 January 2019 (2019-01-31), pages 276 - 282, XP036754546, DOI: 10.1007/s10278-019-00179-2 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113066081B (en) | 2023-07-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | GasHis-Transformer: A multi-scale visual transformer approach for gastric histopathological image detection | |
| CN108022647B (en) | Lung nodule benign and malignant prediction method based on ResNet-inclusion model | |
| CN107886514A (en) | Breast molybdenum target image lump semantic segmentation method based on depth residual error network | |
| CN108648182B (en) | Breast cancer nuclear magnetic resonance image tumor region segmentation method based on molecular subtype | |
| CN111415342A (en) | Attention mechanism fused automatic detection method for pulmonary nodule image of three-dimensional convolutional neural network | |
| Zhong et al. | Cancer image classification based on DenseNet model | |
| CN114565761B (en) | Deep learning-based method for segmenting tumor region of renal clear cell carcinoma pathological image | |
| CN101103924A (en) | Galactophore cancer computer auxiliary diagnosis method based on galactophore X-ray radiography and system thereof | |
| CN108921821A (en) | Method of discrimination based on the LASSO mammary cancer armpit lymph gland transfering state returned | |
| Xu et al. | Using transfer learning on whole slide images to predict tumor mutational burden in bladder cancer patients | |
| CN112215844A (en) | MRI (magnetic resonance imaging) multi-mode image segmentation method and system based on ACU-Net | |
| CN111325750A (en) | Medical image segmentation method based on multi-scale fusion U-shaped chain neural network | |
| Biradar et al. | Lung Cancer detection and classification using 2D convolutional neural network | |
| CN110223331B (en) | Brain MR medical image registration method | |
| Malibari et al. | Artificial intelligence based prostate cancer classification model using biomedical images | |
| Anantharajan et al. | MRI brain tumor detection using deep learning and machine learning approaches | |
| CN117422916A (en) | MR medical image colorectal cancer staging algorithm and system based on weak supervision learning | |
| CN117152179A (en) | Segmentation and classification method for realizing rectal cancer CT image based on U-Net and SENet | |
| Shankara et al. | Detection of lung cancer using convolution neural network | |
| CN117036288A (en) | Tumor subtype diagnosis method for full-slice pathological image | |
| CN113066081B (en) | Mammary tumor molecular subtype detection method based on three-dimensional MRI image | |
| Yang et al. | Ultrasound image-based diagnosis of cirrhosis with an end-to-end deep learning model | |
| CN113870194B (en) | Breast tumor ultrasonic image processing device with fusion of deep layer characteristics and shallow layer LBP characteristics | |
| Li et al. | A dual attention-guided 3D convolution network for automatic segmentation of prostate and tumor | |
| CN114565762A (en) | Weakly supervised liver tumor segmentation based on ROI and split fusion strategy |
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 |