CN112183557A - MSI prediction model construction method based on gastric cancer histopathology image texture features - Google Patents

MSI prediction model construction method based on gastric cancer histopathology image texture features Download PDF

Info

Publication number
CN112183557A
CN112183557A CN202011052554.8A CN202011052554A CN112183557A CN 112183557 A CN112183557 A CN 112183557A CN 202011052554 A CN202011052554 A CN 202011052554A CN 112183557 A CN112183557 A CN 112183557A
Authority
CN
China
Prior art keywords
msi
gastric cancer
texture features
prediction model
image
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
Application number
CN202011052554.8A
Other languages
Chinese (zh)
Inventor
阎婷
安卫超
张楠
王彬
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanxi Medical University
Original Assignee
Shanxi Medical University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shanxi Medical University filed Critical Shanxi Medical University
Priority to CN202011052554.8A priority Critical patent/CN112183557A/en
Publication of CN112183557A publication Critical patent/CN112183557A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
    • G16B40/00—ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V2201/00—Indexing scheme relating to image or video recognition or understanding
    • G06V2201/03—Recognition of patterns in medical or anatomical images

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • General Engineering & Computer Science (AREA)
  • Medical Informatics (AREA)
  • Health & Medical Sciences (AREA)
  • Biophysics (AREA)
  • Bioethics (AREA)
  • Databases & Information Systems (AREA)
  • Epidemiology (AREA)
  • Multimedia (AREA)
  • Public Health (AREA)
  • Software Systems (AREA)
  • Biotechnology (AREA)
  • General Health & Medical Sciences (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Investigating Or Analysing Biological Materials (AREA)

Abstract

The invention discloses an MSI prediction model construction method based on gastric cancer histopathology image texture characteristics, which comprises the following steps of obtaining a histopathology image of a gastric cancer patient, a mark of a focus part and clinical pathology information; extracting texture features of an original image and texture features after wavelet transform aiming at a histopathological image of a gastric cancer patient; performing feature selection on the obtained texture features by using LASSO, and selecting non-zero coefficient features corresponding to a lambda value when 10 times of cross validation errors are minimum to obtain screened texture features; performing linear fitting according to the characteristic value of the selected texture characteristic and the coefficient weight thereof to obtain an MSI label of the gastric cancer patient; combining the MSI label and clinical and pathological information of the patient to construct an MSI prediction model. The method directly predicts the MSI state of the gastric cancer patient based on the easily obtained histopathological image, does not need additional laboratories to carry out gene detection and immunohistochemical analysis, and can realize the detection of the MSI state at lower cost.

Description

MSI prediction model construction method based on gastric cancer histopathology image texture features
Technical Field
The invention relates to the technical field of computer medical image information processing, in particular to an MSI prediction model construction method based on gastric cancer histopathology image texture features.
Background
The traditional MSI detection methods mainly comprise two methods: immunohistochemistry (Immunohistochemistry IHC) and Polymerase Chain Reaction (PCR); IHC reflects MSI state by detecting the expression of mismatch repair gene, PCR carries out genetic analysis by gene mark of specific mononucleotide site; however, both IHC and PCR detection methods need to be performed in a high-capacity tertiary medical center, and high economic and time costs are required, so that the method is difficult to be popularized to every patient in clinical practice. Thus, a large number of potentially immunotherapy-sensitive individuals cannot be provided with timely treatment with immune checkpoint inhibitors, thereby losing the opportunity to control the disease.
Histopathology has been an important tool in cancer diagnosis and prognosis, and its phenotypic information reflects the combined effect of molecular changes on cancer cell behavior and provides a direct visualization tool for assessing disease progression. Histopathologists can classify and grade lesions by assessing histological features such as cell density, tissue architecture, mitotic state, etc. With the progress of microscope imaging technology and computer technology, aided diagnosis models based on pathological images are rapidly developed. Among them, texture analysis becomes an important cancer analysis method, and cancer images are quantitatively analyzed by defining operators for extracting texture attributes. At present, computer aided diagnosis based on statistical texture analysis is a common feature extraction method. For example, the texture features of the tumor region are extracted based on a gray histogram, a gray co-occurrence matrix, and the like. The prediction model based on the texture characteristics of the pathological images of the cancer patients can make up the defects of manual analysis, not only can provide accurate and objective prediction results, but also can reduce the workload of doctors so as to greatly improve the diagnosis efficiency.
The invention provides a gastric cancer MSI prediction method based on histopathological images, which does not need additional laboratories to carry out gene detection and immunohistochemical analysis and can realize the detection of MSI state with lower cost.
Disclosure of Invention
The invention aims to overcome the problems in the prior art and provide an MSI prediction model construction method based on gastric cancer histopathology image texture features, which is used for extracting quantitative image features from a gastric cancer histopathology image aiming at tumor heterogeneity, constructing a prediction label by using Lasso regression, taking the prediction label as an independent prediction factor, combining clinical pathology information of a patient, performing multivariate analysis by logics regression to construct a prediction model, and drawing a nomogram to realize visualization of the prediction model.
In order to achieve the technical purpose and achieve the technical effect, the invention is realized by the following technical scheme:
an MSI prediction model construction method based on gastric cancer histopathology image texture features comprises the following steps:
step 1, acquiring a histopathological image, a mark of a focus part and clinical pathological information of a gastric cancer patient;
step 2, extracting texture features of an original image and texture features after wavelet transformation aiming at the histopathological image of a gastric cancer patient;
step 3, using LASSO to perform feature selection on the texture features obtained in the step 2, and selecting non-zero coefficient features corresponding to the lambda value when the cross validation error is 10 times as small as the minimum to obtain screened texture features;
step 4, performing linear fitting according to the characteristic values of the texture features selected in the step 3 and the coefficient weights thereof to obtain an MSI label of the gastric cancer patient;
and 5, combining the MSI label obtained in the step 4 and clinical and pathological information of the patient to construct an MSI prediction model.
Further, according to the method for constructing the MSI prediction model based on the texture features of the gastric cancer histopathology image, the texture features extracted in step 2 include a First order statistics (First order statistics), a gray level co-occurrence matrix (GLCM), a gray level size area matrix (GLSZM), a Gray Level Run Length Matrix (GLRLM), an adjacent gray level hue difference matrix (NGTDM), and a Gray Level Dependency Matrix (GLDM).
Further, in the method for constructing the MSI prediction model based on the texture features of the histopathology image of gastric cancer as described above, the specific operation of LASSO screening the texture features of the image in step 3 includes:
1) applying LASSO regression to the texture features extracted in the step 2, performing feature selection on high-dimensional data and regularizing, improving prediction accuracy through a penalty estimation function, adding an L1 penalty term to a common linear model, and estimating as follows:
Figure BDA0002709996240000031
where Y is the prediction label, X is the feature vector, λ is the regularization coefficient, λ is the prediction index>0, as the parameter λ becomes larger, most covariate coefficients shrink to zero; beta is ajD is the dimension of the feature matrix;
2) and selecting the model optimal value when the cross validation error is minimum, and screening out the characteristic with the coefficient not being 0 as the texture characteristic most relevant to the MSI state.
Further, in the method for constructing the MSI prediction model based on the texture features of the histopathological image of the gastric cancer, the concrete operation of constructing the MSI tag of the gastric cancer patient in the step 4 includes:
and (3) performing linear fitting according to the texture features screened in the step (3) and the respective coefficient weights and the corresponding feature values, wherein the MSI label calculation formula is as follows:
MSI-tag＝FeatureValvei*αi
wherein alpha isiIs a coefficient of the ith Feature, Feature ValveiIs the patient's ith characteristic value;
and calculating to obtain the MSI label of the gastric cancer patient.
Further, in the method for constructing the MSI prediction model based on the texture features of the histopathological image of gastric cancer as described above, the specific operation of constructing the MSI prediction model in step 5 includes:
and (3) combining clinical pathological information and image scores, constructing a prediction model through logistic regression, and drawing a Nomogram graph to visualize the prediction model.
Further, as described above, the method for constructing the MSI prediction model based on the texture features of the histopathological image of the gastric cancer is a contraction estimation method based on the concept of reducing the feature set, and the LASSO can compress the coefficients of the features and change some regression coefficients to 0, thereby achieving the purpose of feature selection. λ is the regularization parameter, non-zero coefficient feature: the LASSO integrates the feature selection process and the learner training process into a whole, and the feature selection process and the learner training process are completed in the same optimization process, namely, the feature selection is automatically performed in the learner training process. The non-zero coefficient here is the regression coefficient of the linear model obtained by the learner, and the non-zero coefficient feature is a new feature vector obtained after feature selection. Different lambda values will produce different regression models, and as lambda increases, the penalty increases and more feature coefficients are compressed to 0.
The invention has the beneficial effects that:
1. according to the invention, quantitative image features are extracted from a gastric cancer histopathology image aiming at tumor heterogeneity, a prediction label is constructed by using Lasso regression, the prediction label is used as an independent prediction factor, multivariate analysis is carried out by combining clinical pathology information of a patient to construct a prediction model, and a nomogram is drawn to realize the visualization of the prediction model. The texture features of the histopathological image of the gastric cancer patient are extracted, the extracted texture analysis features can highlight the tumor characteristics of the patient, and the constructed prediction model is more effective.
2. The method directly predicts the MSI state of the gastric cancer patient by using a machine learning technology based on an easily obtained histopathology image without additional laboratories for gene detection and immunohistochemical analysis, and can realize the detection of the MSI state at lower cost.
Of course, it is not necessary for any one product that embodies the invention to achieve all of the above advantages simultaneously.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced 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 that other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a ROC plot of MSI signature versus MSI status discrimination in gastric cancer patients;
FIG. 3 is a calibration graph of the prediction model before and after adding the MSI tag, where A is before adding the MSI tag and B is after adding the MSI tag;
FIG. 4 is a graph comparing DCA curves of the prediction model before and after MSI tag addition.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Construction of a prediction model:
referring to fig. 1, a method for constructing a gastric cancer MSI prediction model based on texture features of a histopathological image includes the following steps:
step 1, acquiring a histopathological image, a mark of a focus part and clinical pathological information of a gastric cancer patient;
in the embodiment, the technical scheme provided by the invention is applied to the histopathological image data of the gastric cancer patient. The data set is an open data set obtained from a TCGA database, and 277 cases with uniform staining, clear imaging, complete clinical and pathological information and definite MSI state information are obtained after screening. And marking and segmenting the ROI of the pathological image under the guidance of a professional physician. In order to ensure the balance between positive and negative labels of the data set, the data set is up-sampled to obtain 442 independent samples, according to 3: 1 the data set is randomly divided into a training set and a validation set: 313 samples are shared by the training set, wherein 156 samples are MSI types, 157 samples are MSS types; the verification set has 129 samples, of which 64 are MSI type and 65 are MSS type.
Data information for all patients in this example is shown in table 1:
TABLE 1
Figure BDA0002709996240000051
Figure BDA0002709996240000061
And 2, extracting texture features of the original image and texture features after wavelet transformation aiming at the histopathological image of the gastric cancer patient. Texture features include First order statistics (First order statistics), gray level co-occurrence matrices (GLCM), gray level size region matrices (GLSZM), Gray Level Run Length Matrices (GLRLM), adjacent gray level hue difference matrices (NGTDM), and Gray Level Dependency Matrices (GLDM).
And 3, performing feature selection on the texture features obtained in the step 2 by using LASSO, and selecting the non-zero coefficient features corresponding to the lambda value when the cross validation error is 10 times as small as the minimum to obtain the screened texture features. The specific operation comprises the following steps:
1) applying LASSO regression to the texture features extracted in the step 2, performing feature selection on high-dimensional data and regularizing, improving prediction accuracy through a penalty estimation function, adding an L1 penalty term to a common linear model, and estimating as follows:
Figure BDA0002709996240000062
wherein Y isMeasuring labels, wherein X is a characteristic vector, and lambda is a regularization coefficient>0, as the parameter λ becomes larger, most covariate coefficients shrink to zero; beta is ajD is the dimension of the feature matrix;
2) and selecting the model optimal value when the cross validation error is minimum, and screening out the characteristic with the coefficient not being 0 as the texture characteristic most relevant to the MSI state.
And 4, performing linear fitting according to the characteristic values of the texture features selected in the step 3 and the coefficient weights thereof to obtain the MSI label of the gastric cancer patient.
The specific operation comprises the following steps:
and (3) performing linear fitting according to the texture features screened in the step (3) and the respective coefficient weights and the corresponding feature values, wherein the MSI label calculation formula is as follows:
MSI-tag＝FeatureValvei*αi
wherein alpha isiIs a coefficient of the ith Feature, Feature ValveiIs the patient's ith characteristic value;
and calculating to obtain the MSI label of the gastric cancer patient.
And 5, combining the MSI label obtained in the step 4 and clinical and pathological information of the patient to construct an MSI prediction model. The specific operation comprises the following steps:
and (3) combining clinical pathological information and image scores, constructing a prediction model through logistic regression, and drawing a Nomogram graph to visualize the prediction model.
Evaluation of prediction performance:
as shown in fig. 2, the ability of the constructed MSI prediction signature to distinguish MSI states was first verified by the ROC curve, and it can be seen that the MSI prediction signature constructed in this study can better distinguish MSI states of gastric cancer patients both in the training set and in the verification set.
As shown in fig. 3, the gains of the texture features of the histopathology image to the prediction model are further verified, and the calibration curves of the prediction model before and after adding the MSI label are plotted, so that the calibration curve of the prediction model after adding the MSI label is better in performance.
As shown in fig. 4, to verify the clinical utility of the prediction model, we evaluated the clinical application value of nomograms based on texture features of pathological images by decision curve analysis and quantification of net profit. And in the whole risk threshold interval, the prediction model after the MSI label is added obtains larger net benefit.
In conclusion, the method for constructing the gastric cancer MSI prediction model based on the texture features of the histopathological image has the capability of practical clinical application.
The preferred embodiments of the invention disclosed above are intended to be illustrative only. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise embodiments disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims (6)

1. A construction method of an MSI prediction model based on texture features of a gastric cancer histopathology image is characterized by comprising the following steps:
step 1, acquiring a histopathological image, a mark of a focus part and clinical pathological information of a gastric cancer patient;
step 2, extracting texture features of an original image and texture features after wavelet transformation aiming at the histopathological image of a gastric cancer patient;
step 3, using LASSO to perform feature selection on the texture features obtained in the step 2, and selecting non-zero coefficient features corresponding to the lambda value when the cross validation error is 10 times as small as the minimum to obtain screened texture features;
step 4, performing linear fitting according to the characteristic values of the texture features selected in the step 3 and the coefficient weights thereof to obtain an MSI label of the gastric cancer patient;
and 5, combining the MSI label obtained in the step 4 and clinical and pathological information of the patient to construct an MSI prediction model.
2. The method for constructing the MSI prediction model based on the texture features of the gastric cancer histopathology image according to claim 1, wherein the texture features extracted in the step 2 include a first order statistic, a gray level co-occurrence matrix, a gray level size region matrix, a gray level run length matrix, an adjacent gray level hue difference matrix, and a gray level dependency matrix.
3. The method for constructing the MSI prediction model based on the texture features of the histopathological images of gastric cancer as claimed in claim 1, wherein the LASSO filters the texture features of the images in step 3 by:
1) applying LASSO regression to the texture features extracted in the step 2, performing feature selection on high-dimensional data and regularizing, improving prediction accuracy through a penalty estimation function, adding an L1 penalty term to a common linear model, and estimating as follows:
Figure FDA0002709996230000011
where Y is the prediction label, X is the feature vector, λ is the regularization coefficient, λ is the prediction index>0, as the parameter λ becomes larger, most covariate coefficients shrink to zero; beta is ajD is the dimension of the feature matrix;
2) and selecting the model optimal value when the cross validation error is minimum, and screening out the characteristic with the coefficient not being 0 as the texture characteristic most relevant to the MSI state.
4. The method for constructing the MSI prediction model based on the texture features of the histopathological image of the gastric cancer as claimed in claim 1, wherein the concrete operation of constructing the MSI label of the gastric cancer patient in the step 4 comprises:
and (3) performing linear fitting according to the texture features screened in the step (3) and the respective coefficient weights and the corresponding feature values, wherein the MSI label calculation formula is as follows:
MSI-tag＝FeatureValvei*αi
wherein alpha isiIs a coefficient of the ith Feature, Feature ValveiIs the patient's ith characteristic value;
and calculating to obtain the MSI label of the gastric cancer patient.
5. The method for constructing an MSI prediction model based on texture features of a histopathological image of gastric cancer according to claim 1, wherein the concrete operation of constructing the MSI prediction model in step 5 comprises:
and (3) combining clinical pathological information and image scores, constructing a prediction model through logistic regression, and drawing a Nomogram graph to visualize the prediction model.
6. The method for constructing an MSI prediction model based on texture features of histopathological images of gastric cancer as claimed in claim 1, wherein the LASSO is a contraction estimation method based on a reduced feature set, and LASSO can compress coefficients of features and change some regression coefficients to 0, thereby achieving the purpose of feature selection.
CN202011052554.8A 2020-09-29 2020-09-29 MSI prediction model construction method based on gastric cancer histopathology image texture features Pending CN112183557A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011052554.8A CN112183557A (en) 2020-09-29 2020-09-29 MSI prediction model construction method based on gastric cancer histopathology image texture features

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011052554.8A CN112183557A (en) 2020-09-29 2020-09-29 MSI prediction model construction method based on gastric cancer histopathology image texture features

Publications (1)

Publication Number Publication Date
CN112183557A true CN112183557A (en) 2021-01-05

Family

ID=73946562

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011052554.8A Pending CN112183557A (en) 2020-09-29 2020-09-29 MSI prediction model construction method based on gastric cancer histopathology image texture features

Country Status (1)

Country Link
CN (1) CN112183557A (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113361580A (en) * 2021-05-31 2021-09-07 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） Colorectal cancer mismatch repair function prediction method and device
CN113921079A (en) * 2021-12-06 2022-01-11 四川省肿瘤医院 MSI prediction model construction method based on immune related gene
CN115132327A (en) * 2022-05-25 2022-09-30 中国医学科学院肿瘤医院 Microsatellite instability prediction system, construction method thereof, terminal equipment and medium
CN115294231A (en) * 2022-09-29 2022-11-04 浙江省肿瘤医院 Prediction method of mismatch repair gene defect of gastric cancer
CN115295126A (en) * 2022-09-29 2022-11-04 浙江省肿瘤医院 Model for predicting mismatch repair gene defect of gastric cancer

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108701173A (en) * 2015-10-12 2018-10-23 南托米克斯有限责任公司 System, composition and method for finding the prediction MSI and new epitope sensitive to checkpoint inhibitor
CN109196554A (en) * 2016-05-18 2019-01-11 豪夫迈·罗氏有限公司 Tumour measures of closeness
CN111028223A (en) * 2019-12-11 2020-04-17 大连医科大学附属第一医院 Microsatellite unstable intestinal cancer energy spectrum CT iodine water map image omics feature processing method
CN111340770A (en) * 2020-02-22 2020-06-26 太原理工大学 Method for constructing cancer prognosis model by combining global weighted LBP (local binary pattern) and texture analysis

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108701173A (en) * 2015-10-12 2018-10-23 南托米克斯有限责任公司 System, composition and method for finding the prediction MSI and new epitope sensitive to checkpoint inhibitor
CN109196554A (en) * 2016-05-18 2019-01-11 豪夫迈·罗氏有限公司 Tumour measures of closeness
CN111028223A (en) * 2019-12-11 2020-04-17 大连医科大学附属第一医院 Microsatellite unstable intestinal cancer energy spectrum CT iodine water map image omics feature processing method
CN111340770A (en) * 2020-02-22 2020-06-26 太原理工大学 Method for constructing cancer prognosis model by combining global weighted LBP (local binary pattern) and texture analysis

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
武敬君等: ""能谱CT成像碘(水)图纹理分析预测结直肠癌微卫星不稳定状态"", 《中国医学影像技术》 *

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113361580A (en) * 2021-05-31 2021-09-07 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） Colorectal cancer mismatch repair function prediction method and device
CN113921079A (en) * 2021-12-06 2022-01-11 四川省肿瘤医院 MSI prediction model construction method based on immune related gene
CN115132327A (en) * 2022-05-25 2022-09-30 中国医学科学院肿瘤医院 Microsatellite instability prediction system, construction method thereof, terminal equipment and medium
WO2023226217A1 (en) * 2022-05-25 2023-11-30 中国医学科学院肿瘤医院 Microsatellite instability prediction system and construction method therefor, terminal device, and medium
CN115294231A (en) * 2022-09-29 2022-11-04 浙江省肿瘤医院 Prediction method of mismatch repair gene defect of gastric cancer
CN115295126A (en) * 2022-09-29 2022-11-04 浙江省肿瘤医院 Model for predicting mismatch repair gene defect of gastric cancer
CN115294231B (en) * 2022-09-29 2023-02-03 浙江省肿瘤医院 Method for predicting mismatch repair gene defect of gastric cancer
CN115295126B (en) * 2022-09-29 2023-03-24 浙江省肿瘤医院 Model for predicting mismatch repair gene defect of gastric cancer

Similar Documents

Publication Publication Date Title
CN112183557A (en) MSI prediction model construction method based on gastric cancer histopathology image texture features
Zhu et al. Wsisa: Making survival prediction from whole slide histopathological images
KR102583103B1 (en) Systems and methods for processing electronic images for computational detection methods
CN110021425B (en) Comparison detector, construction method thereof and cervical cancer cell detection method
US20220058839A1 (en) Translation of images of stained biological material
CN111340770B (en) Method for constructing cancer prognosis model by combining global weighted LBP (local binary pattern) and texture analysis
Sheehan et al. Detection and classification of novel renal histologic phenotypes using deep neural networks
CN113706434B (en) Post-processing method for chest enhancement CT image based on deep learning
CN112543934A (en) Method for determining degree of abnormality, corresponding computer readable medium and distributed cancer analysis system
Xu et al. Using transfer learning on whole slide images to predict tumor mutational burden in bladder cancer patients
US11804029B2 (en) Hierarchical constraint (HC)-based method and system for classifying fine-grained graptolite images
CN113393454A (en) Method and device for segmenting pathological target examples in biopsy tissues
CN114445356A (en) Multi-resolution-based full-field pathological section image tumor rapid positioning method
CN116468690B (en) Subtype analysis system of invasive non-mucous lung adenocarcinoma based on deep learning
CN109191452B (en) Peritoneal transfer automatic marking method for abdominal cavity CT image based on active learning
Arar et al. High-quality immunohistochemical stains through computational assay parameter optimization
CN115239947A (en) Wheat stripe rust severity evaluation method and device based on unsupervised learning
CN114267434A (en) Cancer patient tumor image delineation method based on image omics
CN113762395A (en) Pancreatic bile duct type ampulla carcinoma classification model generation method and image classification method
Athanasiadis et al. Segmentation of complementary DNA microarray images by wavelet-based Markov random field model
CN112703531A (en) Generating annotation data for tissue images
CN117133466B (en) Survival prediction method and device based on transcriptomics and image histology
CN111553369B (en) Gralison automatic grading method and device for prostate cancer puncture pathological image
CN116030017B (en) Pathological prognosis analysis method and system based on multi-modal convolutional neural network
WO2023240819A1 (en) Pathological analysis module for thyroid disease

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
RJ01 Rejection of invention patent application after publication

Application publication date: 20210105