CN108898160B - Breast cancer histopathology grading method based on CNN and imaging omics feature fusion - Google Patents
Breast cancer histopathology grading method based on CNN and imaging omics feature fusion Download PDFInfo
- Publication number
- CN108898160B CN108898160B CN201810555876.0A CN201810555876A CN108898160B CN 108898160 B CN108898160 B CN 108898160B CN 201810555876 A CN201810555876 A CN 201810555876A CN 108898160 B CN108898160 B CN 108898160B
- Authority
- CN
- China
- Prior art keywords
- cnn
- model
- image
- breast cancer
- molybdenum target
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of CNN and image classification and identification, in particular to a breast cancer histopathology grading method based on CNN and image omics feature fusion. The invention provides a method for judging the histopathology grade of breast cancer of a molybdenum target image by constructing a CNN model with fused characteristics, utilizing the gray characteristics, texture characteristics and wavelet characteristics extracted from a molybdenum target tumor region, performing characteristic screening by an LASSO logistic regression model, selecting the characteristics with high correlation with the histopathology grade of the breast cancer, performing characteristic fusion on the high-level semantic characteristics extracted by the CNN and the screened image omics characteristics in a fully-connected layer newly added in a network, and fitting to obtain the CNN model with fused characteristics for identifying the histopathology grade of the breast cancer. The method can directly analyze and judge the histopathology grade of the breast cancer of the patient according to the breast molybdenum target image scanned by the patient, and further shortens the judgment time while ensuring the judgment precision.
Description
Technical Field
The invention relates to the technical field of CNN and image classification and identification, in particular to a breast cancer histopathology grading method based on CNN and image omics feature fusion.
Background
Breast cancer, a common cancer in women, is the second most lethal disease in women. The incidence of breast cancer has been rising throughout the world since the end of the 70 s of the 20 th century, and there are few patients who died from breast cancer. The mammary gland molybdenum target X-ray photographic examination technology is the first choice and the simplest and most reliable noninvasive detection means for judging mammary gland diseases at present, has high resolution and is beneficial to early discovery of breast cancer.
In recent years, with the development of large data and high performance computing, CNN (convolutional neural network) has achieved remarkable performance in the field of computer vision, and the recognition rate in natural image classification has exceeded the human recognition level. The CNN extracts image features through multilayer convolution and pooling, and then performs parameter updating through a back propagation algorithm, so that the limitation that the original manual design features are subjected to experience of people is changed.
The histopathological grading (SBR grading) of the breast cancer is mainly evaluated by combining image morphological characteristics of three aspects of mitosis index of cancer cells, difference of mammary gland ducts and abnormal shape of cancer cell nucleuses, the histopathological grading of the breast cancer and the prognosis of a patient have important relation, and the 5-year survival rate of the patient is reduced along with the increase of the histopathological grading in the same clinical stage. The Sequencing Batch Reactor (SBR) classification of the breast cancer is mainly used for observing the cancer cell differentiation condition of pathological sections of a patient under a microscope, and at present, doctors cannot directly perform classification judgment on conventional molybdenum target images.
Disclosure of Invention
Aiming at the problems, the invention provides a breast cancer histopathology grading method based on CNN and image omics feature fusion, which can directly analyze a breast molybdenum target image of a patient, perform feature fusion on a newly added full-connection layer by using an artificially designed image omics feature and an image high-level semantic feature automatically extracted by CNN, train a CNN model after feature fusion so as to obtain the breast cancer histopathology grade of the patient and provide a basis for further disease discrimination and prognosis analysis.
In order to achieve the purpose, the invention adopts the following technical scheme:
the breast cancer histopathology grading method based on CNN and imaging omics feature fusion comprises the following steps:
step 1: extracting a mammary molybdenum target image tumor region, calculating gray scale, texture and wavelet characteristics on the extracted molybdenum target tumor region, and extracting 180-dimensional image omics characteristic vectors through the calculation; making the extracted mammary gland molybdenum target image tumor region into mammary gland tumor region molybdenum target image samples with the same size, and dividing the image samples into a training set, a verification set and a test set;
step 2: performing feature screening on the extracted 180-dimensional image omics feature vector by adopting an LASSO logistic regression model, and performing feature fusion by using the screened image omics features;
and step 3: the method comprises the steps of performing transfer learning by adopting a pre-trained CNN model, training a CNN hierarchical model, adding a new full-connection layer on the original basis of the CNN hierarchical model, performing feature fusion on the output of the CNN hierarchical model before the full-connection layer and the screened image omics features on the new full-connection layer, performing retraining on the basis of the CNN hierarchical model parameters, updating the fused CNN hierarchical model parameters, and adjusting the fused CNN hierarchical model parameters according to the hierarchical effect of the model on a verification set to obtain a feature-fused CNN model for performing breast cancer histopathology classification on a breast molybdenum target image.
Further, after the step 3, the method further comprises the following steps:
and (5) verifying the classification accuracy of the model by using the test set pair to obtain the CNN model with the fused features.
Further, the step 1 comprises:
step 1.1: extracting ROI from a tumor region of a breast molybdenum target image to obtain an ROI image, calculating 14 gray features, 22 texture features and 144 wavelet features of the ROI image, and extracting 180-dimensional image omics feature vectors in total;
step 1.2: expanding the scale of the ROI image by a data enhancement method;
step 1.3: and uniformly scaling the ROI image after data scale expansion to the same size to adapt to the input requirement of the CNN model.
Further, the step 3 comprises:
step 3.1: taking a molybdenum target image sample of a breast tumor area in a training set as the input of a CNN model, carrying out transfer learning on the CNN model pre-trained on an ImageNet natural image data set, and training a CNN hierarchical model;
step 3.2: adding a new full-connection layer on the original basis of the CNN hierarchical model, performing feature fusion on the new full-connection layer by outputting the high-level semantic features of the molybdenum target image in the breast tumor region before the CNN hierarchical model full-connection layer and the omics features screened by the LASSO logistic regression model, performing retraining on the basis of the CNN hierarchical model parameters, updating the fused CNN hierarchical model parameters, and adjusting the fused CNN hierarchical model parameters according to the hierarchical effect of the model on the verification set to obtain the feature-fused CNN model.
Compared with the prior art, the invention has the following beneficial effects:
the invention provides a method for judging the histopathology grade of breast cancer of a molybdenum target image by constructing a CNN model with fused characteristics, utilizing the gray characteristics, texture characteristics and wavelet characteristics extracted from a molybdenum target tumor region, performing characteristic screening by an LASSO logistic regression model, selecting the characteristics with high correlation with the histopathology grade of the breast cancer, performing characteristic fusion on the high-level semantic characteristics extracted by the CNN and the screened image omics characteristics in a fully-connected layer newly added in a network, and fitting to obtain the CNN model with fused characteristics for identifying the histopathology grade of the breast cancer. The method can directly analyze and judge the histopathology grade of the breast cancer of the patient according to the breast molybdenum target image scanned by the patient, and further shortens the judgment time while ensuring the judgment precision.
Drawings
Fig. 1 is a basic flowchart of a breast cancer histopathological grading method based on CNN and proteomics feature fusion according to an embodiment of the present invention.
Fig. 2 is a basic flowchart of a breast cancer histopathological grading method based on CNN and proteomics feature fusion according to another embodiment of the present invention.
Fig. 3 is a molybdenum target image of different projection positions of a breast cancer histopathological grading method based on CNN and imagery omics feature fusion according to an embodiment of the present invention.
Detailed Description
The invention is further illustrated by the following examples in conjunction with the accompanying drawings:
the first embodiment is as follows:
as shown in fig. 1, the breast cancer histopathological grading method based on CNN and proteomics feature fusion of the present invention comprises the following steps:
step S101: extracting a mammary molybdenum target image tumor region, calculating gray scale, texture and wavelet characteristics on the extracted molybdenum target tumor region, and extracting 180-dimensional image omics characteristic vectors through the calculation; and making the extracted mammary gland molybdenum target image tumor region into mammary gland tumor region molybdenum target image samples with the same size, and dividing the image samples into a training set, a verification set and a test set.
Step S102: and (3) performing feature screening on the extracted 180-dimensional image omics feature vector by adopting an LASSO (laser-induced plasticity) logistic regression model, and performing feature fusion by using the screened image omics features.
Step S103: the method comprises the steps of performing transfer learning by adopting a pre-trained CNN model, training a CNN hierarchical model, adding a new full-connection layer on the original basis of the CNN hierarchical model, performing feature fusion on the output of the CNN hierarchical model before the full-connection layer and the screened image omics features on the new full-connection layer, performing retraining on the basis of the CNN hierarchical model parameters to obtain a fused CNN model, adjusting the fused CNN hierarchical model parameters according to the hierarchical effect of the model on a verification set to obtain the feature-fused CNN model, and performing breast cancer histopathology classification on a breast molybdenum target image.
Example two:
as shown in fig. 2, another histopathological grading method for breast cancer based on CNN and proteomics feature fusion of the present invention comprises the following steps:
step S201: extracting a mammary molybdenum target image tumor region, calculating gray scale, texture and wavelet characteristics on the extracted molybdenum target tumor region, and extracting 180-dimensional image omics characteristic vectors through the calculation; and making the extracted mammary gland molybdenum target image tumor region into mammary gland tumor region molybdenum target image samples with the same size, and dividing the image samples into a training set, a verification set and a test set.
The step S201 includes:
step S2011: extracting ROI from a tumor region of a breast molybdenum target image to obtain an ROI image, calculating 14 gray features, 22 texture features and 144 wavelet features of the ROI image, and extracting 180-dimensional image omics feature vectors in total;
the gray features are 14 features of maximum gray value, minimum gray value, mean value, median, variance, kurtosis, energy, entropy, mean absolute variance, skewness, standard deviation, uniformity, gray value domain and root mean square (see: Aerts H J W L, Velazquez E R, Leijenaar R T H, et al, decoding of graphics by binary imaging using a quantitative radio approach [ J ]. Nature communications,2014,5: 4006);
the texture features are 9-dimensional features derived based on the Gray Level co-occurrence matrix, namely energy, contrast, entropy, homogeneity, correlation, Variance, and mean, difference, autocorrelation (see Weszka J S, Dyer C R, Rosenfeld A.A comparative study of texture measures for tertiary in classification [ J ]. IEEE transactions on Systems, Man, and Cybernetics,1976(4):269-285) and 13-dimensional features derived based on the Gray Level Run matrix, namely Short Run Emphasis, Long Run Emphasis, Gray-Level noise, Run probability, Low Gray-Level Run Emsis, High-Level Run, Run-Level noise, Run-Level noise, Run-Level noise, Run-Level, Run-Level noise, Run-Level, Run-Level, Run-Level, Level-Level, Level-Level, Level-Level, Level-Level, and Level-, Run-Length Variance (see, Galloway M. texture analysis using grey level Run Length hs [ J ]. NASA STI/Recon Technical Report N,1974, 75; Chu A, Sehgal C M, Greenleaf J F. use of grade value distribution of Run Length hs for texture analysis [ J ]. Pattern Recognition Letters,1990,11(6): 415-;
the wavelet features are 144 features, which are respectively calculated on 4 wavelet components.
Step S2012: expanding the scale of the ROI image by a data enhancement method; as an implementable way, the scale of the ROI image can be extended by data enhancement methods of random translation, rotation, flipping, and multi-scale scaling;
step S2013: and uniformly scaling the ROI image after data scale expansion to the same size to adapt to the input requirement of the CNN model.
Step S202: and (3) performing feature screening on the extracted 180-dimensional imaging group feature vector by adopting an LASSO logistic regression model, selecting features with high correlation with the histopathology grade of the breast cancer, and performing feature fusion by using the screened imaging group features.
LASSO regression is a method of adding L1 regularization terms on the basis of least square fitting to improve the accuracy of a linear regression model, and the penalty function of the LASSO regression is the absolute value of the regression coefficient, so that some parameter estimation results are equal to zero, and feature selection is facilitated. Histopathological grading is a binary classification problem, and Logistic regression analysis is a generalized linear model commonly used for binary classification or one-to-many classification, which normalizes the response of simple linear regression to 0 and 1, so that linear regression in the LASSO regression model can be replaced by Logistic regression to pick out the features of binary classification. The objective function of the LASSO logistic regression optimization is as follows:
where n is the number of samples, XiIs a raw data of size m × n, i.e. each sample has m features, yiIs the response value corresponding to each sample, ω is the linear regression coefficient, b is the cutoff value of the linear regression, and λ is the non-negative regularization parameter used to control the sparsity of the regression coefficient. And inputting the extracted image omics characteristics into an LASSO logistic regression model for image omics characteristic screening.
Step S203: the method comprises the steps of performing transfer learning by adopting a pre-trained CNN model, training a CNN hierarchical model, adding a new full-connection layer on the original basis of the CNN hierarchical model, performing feature fusion on the output of the CNN hierarchical model before the full-connection layer and the screened image omics features on the new full-connection layer, performing retraining on the basis of the CNN hierarchical model parameters, updating the fused CNN hierarchical model parameters, and adjusting the fused CNN hierarchical model parameters according to the hierarchical effect of the model on a verification set to obtain a feature-fused CNN model for performing breast cancer histopathology classification on a breast molybdenum target image.
The step S203 includes:
step S2031: taking a molybdenum target image sample of a breast tumor area in a training set as the input of a CNN model, carrying out transfer learning on the CNN model pre-trained on an ImageNet natural image data set, and training a CNN hierarchical model;
step S2032: adding a new full-connection layer on the original basis of the CNN hierarchical model, performing feature fusion on the new full-connection layer by outputting the high-level semantic features of the molybdenum target image in the breast tumor region before the CNN hierarchical model full-connection layer and the omics features screened by the LASSO logistic regression model, performing retraining on the basis of the CNN hierarchical model parameters, updating the fused CNN hierarchical model parameters, and adjusting the fused CNN hierarchical model parameters according to the hierarchical effect of the model on the verification set to obtain the feature-fused CNN model.
Step S204: and verifying the grading accuracy of the model by using the test set on the obtained characteristic-fused CNN model.
As an example, the breast molybdenum target image dataset used comprises 204 cases, each case including an axial (CC) image and a lateral oblique (MLO) image, as shown in fig. 3, where part (a) in fig. 3 is an axial molybdenum target image and part (b) in fig. 3 is a diagonal molybdenum target image. Molybdenum target image storage uses the standard DICOM format with resolution (width x height) of both 3328 x 4096 and 2560 x 3328. The tumor area in all molybdenum target images was delineated by hospital professional radiologists and all cases were fitted with accurate diagnosis in hospital pathology to determine their grade of pathology. The classification performance was quantitatively evaluated by using classification accuracy and AUC values through testing the collected mammary molybdenum target image data sets by different mammary molybdenum target image histopathological grading algorithms, and the results are shown in table 1. Compared with GoogleLeNet (see Szegedy C, Liu W, Jia Y, et al. going stripper with contents [ C ]. IEEE Conference on Computer Vision and Pattern Recognition 2015:1-9), the breast cancer histopathology grading method based on CNN and image omics feature fusion provided by the embodiment of the invention has the advantages that the classification effect is remarkably improved, the classification accuracy reaches 0.7500, and the AUC value reaches 0.8051.
TABLE 1 Classification Performance of molybdenum target breast cancer image pathology grading Algorithm
| Classification algorithm | Accuracy of classification | AUC |
| GoogLeNet | 0.7031 | 0.7049 |
| Random forest | 0.6029 | 0.6618 |
| Feature fusion algorithm | 0.7500 | 0.8051 |
The above shows only the preferred embodiments of the present invention, and it should be noted that it is obvious to those skilled in the art that various modifications and improvements can be made without departing from the principle of the present invention, and these modifications and improvements should also be considered as the protection scope of the present invention.
Claims (3)
1. The breast cancer histopathology grading method based on CNN and imaging omics feature fusion is characterized by comprising the following steps of:
step 1: extracting a mammary molybdenum target image tumor region, calculating gray scale, texture and wavelet characteristics on the extracted molybdenum target tumor region, and extracting 180-dimensional image omics characteristic vectors through the calculation; making the extracted mammary gland molybdenum target image tumor region into mammary gland tumor region molybdenum target image samples with the same size, and dividing the image samples into a training set, a verification set and a test set;
step 2: performing feature screening on the extracted 180-dimensional image omics feature vector by adopting an LASSO logistic regression model, and performing feature fusion by using the screened image omics features;
and step 3: adopting a pre-trained CNN model to perform migration learning, training a CNN hierarchical model, adding a new full-connection layer on the original basis of the CNN hierarchical model, performing feature fusion on the output before the CNN hierarchical model full-connection layer and the screened image omics features on the new full-connection layer, performing retraining on the basis of the CNN hierarchical model parameters, updating the fused CNN hierarchical model parameters, and adjusting the fused CNN hierarchical model parameters according to the hierarchical effect of the model on a verification set to obtain a feature-fused CNN model for performing breast cancer histopathological classification on a breast molybdenum target image;
the step 3 comprises the following steps:
step 3.1: taking a molybdenum target image sample of a breast tumor area in a training set as the input of a CNN model, carrying out transfer learning on the CNN model pre-trained on an ImageNet natural image data set, and training a CNN hierarchical model;
step 3.2: adding a new full-connection layer on the original basis of the CNN hierarchical model, performing feature fusion on the new full-connection layer by outputting the high-level semantic features of the molybdenum target image in the breast tumor region before the CNN hierarchical model full-connection layer and the omics features screened by the LASSO logistic regression model, performing retraining on the basis of the CNN hierarchical model parameters, updating the fused CNN hierarchical model parameters, and adjusting the fused CNN hierarchical model parameters according to the hierarchical effect of the model on the verification set to obtain the feature-fused CNN model.
2. The method for histopathological staging of breast cancer based on fusion of CNN and omics features according to claim 1, further comprising, after said step 3:
and (5) verifying the classification accuracy of the model by using the test set pair to obtain the CNN model with the fused features.
3. The method for histopathological staging of breast cancer based on fusion of CNN and proteomic features according to claim 1, wherein said step 1 comprises:
step 1.1: extracting ROI from a tumor region of a breast molybdenum target image to obtain an ROI image, calculating 14 gray features, 22 texture features and 144 wavelet features of the ROI image, and extracting 180-dimensional image omics feature vectors in total;
step 1.2: expanding the scale of the ROI image by a data enhancement method;
step 1.3: and uniformly scaling the ROI image after data scale expansion to the same size to adapt to the input requirement of the CNN model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810555876.0A CN108898160B (en) | 2018-06-01 | 2018-06-01 | Breast cancer histopathology grading method based on CNN and imaging omics feature fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810555876.0A CN108898160B (en) | 2018-06-01 | 2018-06-01 | Breast cancer histopathology grading method based on CNN and imaging omics feature fusion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108898160A CN108898160A (en) | 2018-11-27 |
| CN108898160B true CN108898160B (en) | 2022-04-08 |
Family
ID=64343881
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810555876.0A Active CN108898160B (en) | 2018-06-01 | 2018-06-01 | Breast cancer histopathology grading method based on CNN and imaging omics feature fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108898160B (en) |
Families Citing this family (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109583447A (en) * | 2018-12-11 | 2019-04-05 | 山东省肿瘤防治研究院（山东省肿瘤医院） | A kind of image group credit analysis model building method and analysis method |
| CN109816632B (en) * | 2018-12-25 | 2020-12-25 | 东软集团股份有限公司 | Brain image processing method and device, readable storage medium and electronic equipment |
| CN109919187B (en) * | 2019-01-28 | 2021-02-12 | 浙江工商大学 | Method for classifying thyroid follicular picture by using bagging fine tuning CNN |
| CN109948667A (en) * | 2019-03-01 | 2019-06-28 | 桂林电子科技大学 | Image classification method and device for the prediction of correct neck cancer far-end transfer |
| CN109993740A (en) * | 2019-04-02 | 2019-07-09 | 广州中医药大学(广州中医药研究院) | A kind of cell cancer image grading method and device |
| CN110175978B (en) * | 2019-04-02 | 2021-05-11 | 南方医科大学南方医院 | Liver cancer imaging omics data processing method, system, device and storage medium |
| CN110533683B (en) * | 2019-08-30 | 2022-04-29 | 东南大学 | Image omics analysis method fusing traditional features and depth features |
| CN110689960B (en) * | 2019-09-24 | 2022-08-09 | 广州大学 | Intelligent breast cancer distinguishing method, device, equipment and storage medium |
| CN111242174B (en) * | 2019-12-31 | 2022-09-23 | 浙江大学 | Liver cancer image feature extraction and pathological classification method based on imaging omics |
| CN111415741B (en) * | 2020-03-05 | 2023-09-26 | 北京深睿博联科技有限责任公司 | Mammary gland X-ray image classification model training method based on implicit apparent learning |
| CN111583320B (en) * | 2020-03-17 | 2023-04-07 | 哈尔滨医科大学 | Breast cancer ultrasonic image typing method and system fusing deep convolutional network and image omics characteristics and storage medium |
| CN111739033A (en) * | 2020-06-22 | 2020-10-02 | 中山大学肿瘤防治中心（中山大学附属肿瘤医院、中山大学肿瘤研究所） | Method for establishing breast molybdenum target and MR image omics model based on machine learning |
| CN111914952B (en) * | 2020-08-21 | 2024-03-08 | 山东第一医科大学附属省立医院(山东省立医院) | AD characteristic parameter screening method and system based on deep neural network |
| CN112365440A (en) * | 2020-10-10 | 2021-02-12 | 天津大学 | Yoov 3-based mastopathy image mitosis detection device and method |
| CN113436158B (en) * | 2021-06-21 | 2023-04-18 | 遂宁市中心医院 | Liver lump auxiliary identification method based on deep learning |
| CN114171197B (en) * | 2021-11-12 | 2022-10-04 | 东莞市人民医院 | Breast cancer HER2 state prediction method and related equipment |
| WO2023115333A1 (en) * | 2021-12-21 | 2023-06-29 | 深圳先进技术研究院 | Expert knowledge constraint-based deep learning method for brain glioma grading |
| CN114360718B (en) * | 2022-03-10 | 2022-07-12 | 之江实验室 | Feature fitting-based PET/CT automatic lung cancer diagnosis and classification system and construction method |
| CN114511564B (en) * | 2022-04-19 | 2023-01-24 | 天津市肿瘤医院（天津医科大学肿瘤医院） | Image analysis method for breast cancer residual tumor load based on DCE-MRI |
| CN116883995B (en) * | 2023-07-07 | 2024-07-16 | 广东食品药品职业学院 | Identification system of breast cancer molecular subtype |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106096654A (en) * | 2016-06-13 | 2016-11-09 | 南京信息工程大学 | A kind of cell atypia automatic grading method tactful based on degree of depth study and combination |
| CN106326931A (en) * | 2016-08-25 | 2017-01-11 | 南京信息工程大学 | Mammary gland molybdenum target image automatic classification method based on deep learning |
| CN106599883A (en) * | 2017-03-08 | 2017-04-26 | 王华锋 | Face recognition method capable of extracting multi-level image semantics based on CNN (convolutional neural network) |
| CN106683081A (en) * | 2016-12-17 | 2017-05-17 | 复旦大学 | Brain glioma molecular marker nondestructive prediction method and prediction system based on radiomics |
| CN106682435A (en) * | 2016-12-31 | 2017-05-17 | 西安百利信息科技有限公司 | System and method for automatically detecting lesions in medical image through multi-model fusion |
| CN106780448A (en) * | 2016-12-05 | 2017-05-31 | 清华大学 | A kind of pernicious sorting technique of ultrasonic Benign Thyroid Nodules based on transfer learning Yu Fusion Features |
| CN107330263A (en) * | 2017-06-26 | 2017-11-07 | 成都知识视觉科技有限公司 | A kind of method of area of computer aided breast invasive ductal carcinoma histological grading |
| CN108062753A (en) * | 2017-12-29 | 2018-05-22 | 重庆理工大学 | The adaptive brain tumor semantic segmentation method in unsupervised domain based on depth confrontation study |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10068171B2 (en) * | 2015-11-12 | 2018-09-04 | Conduent Business Services, Llc | Multi-layer fusion in a convolutional neural network for image classification |
-
2018
- 2018-06-01 CN CN201810555876.0A patent/CN108898160B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106096654A (en) * | 2016-06-13 | 2016-11-09 | 南京信息工程大学 | A kind of cell atypia automatic grading method tactful based on degree of depth study and combination |
| CN106326931A (en) * | 2016-08-25 | 2017-01-11 | 南京信息工程大学 | Mammary gland molybdenum target image automatic classification method based on deep learning |
| CN106780448A (en) * | 2016-12-05 | 2017-05-31 | 清华大学 | A kind of pernicious sorting technique of ultrasonic Benign Thyroid Nodules based on transfer learning Yu Fusion Features |
| CN106683081A (en) * | 2016-12-17 | 2017-05-17 | 复旦大学 | Brain glioma molecular marker nondestructive prediction method and prediction system based on radiomics |
| CN106682435A (en) * | 2016-12-31 | 2017-05-17 | 西安百利信息科技有限公司 | System and method for automatically detecting lesions in medical image through multi-model fusion |
| CN106599883A (en) * | 2017-03-08 | 2017-04-26 | 王华锋 | Face recognition method capable of extracting multi-level image semantics based on CNN (convolutional neural network) |
| CN107330263A (en) * | 2017-06-26 | 2017-11-07 | 成都知识视觉科技有限公司 | A kind of method of area of computer aided breast invasive ductal carcinoma histological grading |
| CN108062753A (en) * | 2017-12-29 | 2018-05-22 | 重庆理工大学 | The adaptive brain tumor semantic segmentation method in unsupervised domain based on depth confrontation study |
Non-Patent Citations (5)
| Title |
|---|
| Breast cancer histopathological image classification using Convolutional Neural Networks;Fabio Alexandre Spanhol 等;《2016 International Joint Conference on Neural Networks》;20161103;第2560-2567页 * |
| Development and Validation of a Radiomics Nomogram for Preoperative Prediction of Lymph Node Metastasis in Colorectal Cancer;Yan-qi Huang 等;《Journal of Clinical Oncology》;20160620;第34卷(第18期);第2157-2164页 * |
| Mitosis detection in breast cancer pathology images by combining handcrafted and convolutional neural network features;Haibo Wang 等;《Journal of Medical Imaging》;20141010;第1卷(第3期);第1-9页 * |
| 基于多特征描述的乳腺癌肿瘤病理自动分级;龚磊 等;《计算机应用》;20151210;第35卷(第12期);第3570-3575,3580页 * |
| 基于深度卷积网络和结合策略的乳腺组织病理图像细胞核异型性自动评分;周超 等;《中国生物医学工程学报》;20170630;第36卷(第3期);摘要，第1-4节 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108898160A (en) | 2018-11-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108898160B (en) | Breast cancer histopathology grading method based on CNN and imaging omics feature fusion | |
| CN107886514B (en) | Mammary gland molybdenum target image lump semantic segmentation method based on depth residual error network | |
| Kallenberg et al. | Unsupervised deep learning applied to breast density segmentation and mammographic risk scoring | |
| Sun et al. | Automatic feature learning using multichannel ROI based on deep structured algorithms for computerized lung cancer diagnosis | |
| Wang et al. | An improved dice loss for pneumothorax segmentation by mining the information of negative areas | |
| CN109949276B (en) | Lymph node detection method for improving SegNet segmentation network | |
| US20170249739A1 (en) | Computer analysis of mammograms | |
| CN109447998B (en) | Automatic segmentation method based on PCANet deep learning model | |
| Pan et al. | Mitosis detection techniques in H&E stained breast cancer pathological images: A comprehensive review | |
| CN104933709A (en) | Automatic random-walk CT lung parenchyma image segmentation method based on prior information | |
| CN112163599B (en) | Image classification method based on multi-scale and multi-level fusion | |
| CN112215807A (en) | Cell image automatic classification method and system based on deep learning | |
| CN112434172A (en) | Pathological image prognosis feature weight calculation method and system | |
| Soleymanifard et al. | Multi-stage glioma segmentation for tumour grade classification based on multiscale fuzzy C-means | |
| CN113628197A (en) | Weakly supervised full-section histopathology image classification method based on contrast learning | |
| CN112348059A (en) | Deep learning-based method and system for classifying multiple dyeing pathological images | |
| Nateghi et al. | Automatic detection of mitosis cell in breast cancer histopathology images using genetic algorithm | |
| Cao et al. | An automatic breast cancer grading method in histopathological images based on pixel-, object-, and semantic-level features | |
| CN115100467A (en) | Pathological full-slice image classification method based on nuclear attention network | |
| CN115147600A (en) | GBM multi-mode MR image segmentation method based on classifier weight converter | |
| Jaffar et al. | Ensemble classification of pulmonary nodules using gradient intensity feature descriptor and differential evolution | |
| Debnath et al. | Brain tumour segmentation using memory based learning method | |
| CN117036288A (en) | Tumor subtype diagnosis method for full-slice pathological image | |
| CN115471701A (en) | Lung adenocarcinoma histology subtype classification method based on deep learning and transfer learning | |
| Li et al. | Automated classification of solitary pulmonary nodules using convolutional neural network based on transfer learning 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 |