CN117275578B - Method for constructing multi-mode prediction model of lung cancer lymph node metastasis - Google Patents
Method for constructing multi-mode prediction model of lung cancer lymph node metastasis Download PDFInfo
- Publication number
- CN117275578B CN117275578B CN202311524786.2A CN202311524786A CN117275578B CN 117275578 B CN117275578 B CN 117275578B CN 202311524786 A CN202311524786 A CN 202311524786A CN 117275578 B CN117275578 B CN 117275578B
- Authority
- CN
- China
- Prior art keywords
- prediction model
- macromage
- distmin
- lung cancer
- lymph node
- 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
- 208000007433 Lymphatic Metastasis Diseases 0.000 title claims abstract description 57
- 208000020816 lung neoplasm Diseases 0.000 title claims abstract description 47
- 206010058467 Lung neoplasm malignant Diseases 0.000 title claims abstract description 46
- 201000005202 lung cancer Diseases 0.000 title claims abstract description 46
- 238000000034 method Methods 0.000 title claims abstract description 23
- 206010064571 Gene mutation Diseases 0.000 claims abstract description 43
- 230000035945 sensitivity Effects 0.000 claims abstract description 12
- 238000003125 immunofluorescent labeling Methods 0.000 claims abstract description 4
- 210000004027 cell Anatomy 0.000 claims description 27
- 238000012706 support-vector machine Methods 0.000 claims description 24
- 206010028980 Neoplasm Diseases 0.000 claims description 16
- 238000010276 construction Methods 0.000 claims description 16
- 210000002540 macrophage Anatomy 0.000 claims description 16
- 210000001519 tissue Anatomy 0.000 claims description 14
- 102100036011 T-cell surface glycoprotein CD4 Human genes 0.000 claims description 8
- 239000000427 antigen Substances 0.000 claims description 8
- 108091007433 antigens Proteins 0.000 claims description 8
- 102000036639 antigens Human genes 0.000 claims description 8
- 210000002919 epithelial cell Anatomy 0.000 claims description 8
- 108090000623 proteins and genes Proteins 0.000 claims description 8
- 101000716102 Homo sapiens T-cell surface glycoprotein CD4 Proteins 0.000 claims description 7
- 102100034540 Adenomatous polyposis coli protein Human genes 0.000 claims description 6
- 102100037700 DNA mismatch repair protein Msh3 Human genes 0.000 claims description 6
- 102100029458 Glutamate receptor ionotropic, NMDA 2A Human genes 0.000 claims description 6
- 102100021866 Hepatocyte growth factor Human genes 0.000 claims description 6
- 102100027755 Histone-lysine N-methyltransferase 2C Human genes 0.000 claims description 6
- 101000924577 Homo sapiens Adenomatous polyposis coli protein Proteins 0.000 claims description 6
- 101001027762 Homo sapiens DNA mismatch repair protein Msh3 Proteins 0.000 claims description 6
- 101001125242 Homo sapiens Glutamate receptor ionotropic, NMDA 2A Proteins 0.000 claims description 6
- 101000898034 Homo sapiens Hepatocyte growth factor Proteins 0.000 claims description 6
- 101001008892 Homo sapiens Histone-lysine N-methyltransferase 2C Proteins 0.000 claims description 6
- 101001076408 Homo sapiens Interleukin-6 Proteins 0.000 claims description 6
- 101000721646 Homo sapiens Phosphatidylinositol 3-kinase C2 domain-containing subunit gamma Proteins 0.000 claims description 6
- 101000735456 Homo sapiens Protein mono-ADP-ribosyltransferase PARP3 Proteins 0.000 claims description 6
- 101000868152 Homo sapiens Son of sevenless homolog 1 Proteins 0.000 claims description 6
- 101000669447 Homo sapiens Toll-like receptor 4 Proteins 0.000 claims description 6
- 102000001759 Notch1 Receptor Human genes 0.000 claims description 6
- 108010029755 Notch1 Receptor Proteins 0.000 claims description 6
- 102100025063 Phosphatidylinositol 3-kinase C2 domain-containing subunit gamma Human genes 0.000 claims description 6
- 102100034935 Protein mono-ADP-ribosyltransferase PARP3 Human genes 0.000 claims description 6
- 102100039360 Toll-like receptor 4 Human genes 0.000 claims description 6
- 230000035772 mutation Effects 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 5
- 210000003855 cell nucleus Anatomy 0.000 claims description 5
- 108010074708 B7-H1 Antigen Proteins 0.000 claims description 4
- 101100463133 Caenorhabditis elegans pdl-1 gene Proteins 0.000 claims description 4
- 102100027581 Forkhead box protein P3 Human genes 0.000 claims description 4
- 101000861452 Homo sapiens Forkhead box protein P3 Proteins 0.000 claims description 4
- 101000934372 Homo sapiens Macrosialin Proteins 0.000 claims description 4
- 101000946843 Homo sapiens T-cell surface glycoprotein CD8 alpha chain Proteins 0.000 claims description 4
- 102100025136 Macrosialin Human genes 0.000 claims description 4
- 102100034922 T-cell surface glycoprotein CD8 alpha chain Human genes 0.000 claims description 4
- 238000002790 cross-validation Methods 0.000 claims description 4
- 238000007477 logistic regression Methods 0.000 claims description 4
- 230000003595 spectral effect Effects 0.000 claims description 4
- 101710160107 Outer membrane protein A Proteins 0.000 claims description 3
- 238000001228 spectrum Methods 0.000 claims description 3
- 102100023990 60S ribosomal protein L17 Human genes 0.000 claims description 2
- 102100024222 B-lymphocyte antigen CD19 Human genes 0.000 claims description 2
- 102000017420 CD3 protein, epsilon/gamma/delta subunit Human genes 0.000 claims description 2
- 108050005493 CD3 protein, epsilon/gamma/delta subunit Proteins 0.000 claims description 2
- 102000012422 Collagen Type I Human genes 0.000 claims description 2
- 108010022452 Collagen Type I Proteins 0.000 claims description 2
- 101000980825 Homo sapiens B-lymphocyte antigen CD19 Proteins 0.000 claims description 2
- 101001046686 Homo sapiens Integrin alpha-M Proteins 0.000 claims description 2
- 101000576894 Homo sapiens Macrophage mannose receptor 1 Proteins 0.000 claims description 2
- 101000971513 Homo sapiens Natural killer cells antigen CD94 Proteins 0.000 claims description 2
- 102100022338 Integrin alpha-M Human genes 0.000 claims description 2
- 102100022297 Integrin alpha-X Human genes 0.000 claims description 2
- 206010025067 Lung carcinoma cell type unspecified stage I Diseases 0.000 claims description 2
- 102100025354 Macrophage mannose receptor 1 Human genes 0.000 claims description 2
- 102100021462 Natural killer cells antigen CD94 Human genes 0.000 claims description 2
- 102100024616 Platelet endothelial cell adhesion molecule Human genes 0.000 claims description 2
- 101710089372 Programmed cell death protein 1 Proteins 0.000 claims description 2
- 238000012216 screening Methods 0.000 claims description 2
- 102000008096 B7-H1 Antigen Human genes 0.000 claims 1
- 238000003745 diagnosis Methods 0.000 abstract description 3
- 201000010099 disease Diseases 0.000 abstract description 3
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 abstract description 3
- 238000007481 next generation sequencing Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 210000004940 nucleus Anatomy 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 4
- 210000001165 lymph node Anatomy 0.000 description 4
- 102100024216 Programmed cell death 1 ligand 1 Human genes 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 230000036039 immunity Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 210000000038 chest Anatomy 0.000 description 2
- 238000002591 computed tomography Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 238000013188 needle biopsy Methods 0.000 description 2
- 208000002154 non-small cell lung carcinoma Diseases 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- 208000029729 tumor suppressor gene on chromosome 11 Diseases 0.000 description 2
- 238000010200 validation analysis Methods 0.000 description 2
- 208000010507 Adenocarcinoma of Lung Diseases 0.000 description 1
- 101100263837 Bovine ephemeral fever virus (strain BB7721) beta gene Proteins 0.000 description 1
- 208000005623 Carcinogenesis Diseases 0.000 description 1
- 108091006146 Channels Proteins 0.000 description 1
- 101100316840 Enterobacteria phage P4 Beta gene Proteins 0.000 description 1
- 206010027476 Metastases Diseases 0.000 description 1
- -1 PANCK Proteins 0.000 description 1
- 210000001744 T-lymphocyte Anatomy 0.000 description 1
- 206010064390 Tumour invasion Diseases 0.000 description 1
- 238000009098 adjuvant therapy Methods 0.000 description 1
- 210000000621 bronchi Anatomy 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000036952 cancer formation Effects 0.000 description 1
- 230000009400 cancer invasion Effects 0.000 description 1
- 231100000504 carcinogenesis Toxicity 0.000 description 1
- 239000003153 chemical reaction reagent Substances 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000010166 immunofluorescence Methods 0.000 description 1
- 230000008611 intercellular interaction Effects 0.000 description 1
- 201000005249 lung adenocarcinoma Diseases 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000009401 metastasis Effects 0.000 description 1
- 230000002980 postoperative effect Effects 0.000 description 1
- 238000004393 prognosis Methods 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000002560 therapeutic procedure Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 210000004881 tumor cell Anatomy 0.000 description 1
- 230000004614 tumor growth Effects 0.000 description 1
- 238000002604 ultrasonography Methods 0.000 description 1
Classifications
-
- 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
- G16B20/00—ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
- G16B20/50—Mutagenesis
-
- 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
- G16B20/00—ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
-
- 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
- G16B40/10—Signal processing, e.g. from mass spectrometry [MS] or from PCR
Landscapes
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Biotechnology (AREA)
- Evolutionary Biology (AREA)
- Biophysics (AREA)
- Theoretical Computer Science (AREA)
- Genetics & Genomics (AREA)
- Proteomics, Peptides & Aminoacids (AREA)
- Analytical Chemistry (AREA)
- Chemical & Material Sciences (AREA)
- Signal Processing (AREA)
- Artificial Intelligence (AREA)
- Bioethics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Epidemiology (AREA)
- Evolutionary Computation (AREA)
- Public Health (AREA)
- Software Systems (AREA)
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
Abstract
The invention belongs to the field of disease diagnosis and risk prediction, and particularly relates to a method for constructing a multi-mode prediction model of lung cancer lymph node metastasis by combining gene mutation characteristics and multiple immunofluorescence staining mIF image characteristics and a prediction model system. Compared with a single-mode prediction model based on gene mutation characteristics, the lung cancer lymph node metastasis multi-mode prediction model system constructed by the comprehensive gene mutation characteristics and the mIF image characteristics has higher accuracy, specificity, sensitivity and robustness, and well solves the problems of inaccurate or insufficient sensitivity of the lung cancer lymph node metastasis prediction result in the clinic at present.
Description
Technical Field
The invention belongs to the field of disease diagnosis and risk prediction, and particularly relates to a method for constructing a multi-mode prediction model for lung cancer lymph node metastasis by combining gene mutation characteristics and mIF image characteristics and a prediction model system.
Background
The accurate stage of lung cancer contributes to the selection and prognosis evaluation of treatment schemes, and the most common lung cancer stage method at present is a TNM tumor stage system deduced by WHO, which divides lung cancer into four stages according to three dimensions of primary tumor invasion degree (T), lymph node metastasis (N) and distant metastasis (M), wherein stage I, II and stage IIIa are early lung cancers. N stage is closely related to the final stage of lung cancer, and greatly influences postoperative adjuvant treatment, recurrence and total survival of patients, so that the judgment of the existence of lymph node metastasis has important clinical significance.
Although non-invasive imaging examinations such as Computed Tomography (CT) are commonly used in clinic at present to detect lymph node metastasis, accurate lymph node stage diagnosis still cannot be obtained. Although invasive examinations such as mediastinal speculum and needle biopsy under ultrasound guidance of bronchi have higher accuracy, they are usually performed only for positively indexed lymph nodes and cannot cover all lymph node sites. Predictive models based on patient clinical information have also been reported, but are too broad to result in insufficient accuracy. Thus, there is a need for accurate and comprehensive means for predicting lymph node metastasis.
Currently, the mainstream view considers that the gene mutation is a main cause of tumorigenesis, and detecting the gene mutation in lung cancer is beneficial to prompting the lymph node metastasis of patients. Second, tumor immune microenvironment (Tumor Immune Micro-environment, TIME) is also an important factor affecting tumor growth, whether tumor cells are capable of undergoing lymph node metastasis has a close relationship with TIME, and evaluation of TIME is beneficial for predicting lymph node metastasis in patients. Thus, if the detection of gene mutations in lung cancer and the detection of tumor immune microenvironment could be combined, the prediction of lung cancer lymph node metastasis would be greatly improved.
Based on the problem that specificity and sensitivity of a prediction model of lymph node metastasis by noninvasive imaging and clinical information are insufficient, and invasive examination based on needle biopsy cannot cover all lymph nodes, the inventor combines two technologies of next generation sequencing (Next Generation Sequencing, NGS) and multiple immunofluorescence (multiple Immunofluence, mIF) to detect gene mutation and tumor immunity microenvironment respectively, so that a multi-mode prediction model of lung cancer lymph node metastasis is constructed.
There is no report on related research results in the prior art.
Disclosure of Invention
The invention aims to fill the blank of the prior art, solve the defects of the lung cancer lymph node metastasis prediction model clinically used at present and in the research and development stage, and provide a lung cancer lymph node metastasis multi-mode prediction model construction method and a prediction model system integrating gene mutation characteristics and mIF image characteristics.
In order to achieve the purpose of the invention, the invention adopts the following technical scheme:
in a first aspect, the present invention provides a method for constructing a multi-modal predictive model of lung cancer lymph node metastasis that combines a gene mutation signature and a multiplex immunofluorescent staining (mIF) image signature, the predictive model constructing method comprising the steps of:
step 1: obtaining the mutation characteristics of the genes: obtaining one or more gene mutation characteristics with larger absolute value of coefficient Coef in the classifier;
step 2: acquiring mIF image characteristics:
step 2.1: subjecting a lung cancer tissue section of a patient to mIF using corresponding antibodies to one or more lung cancer cell surface antigens;
step 2.2: shooting detailed images of the dyed tissue slices, analyzing by using software, and obtaining the distribution of various single cells on the slices through cluster analysis;
step 2.3: assessing the effect of Tumor Immune Microenvironment (TIME) on lymph node metastasis in a lung cancer patient;
step 2.4: obtaining one or more mIF image features with larger absolute values of coefficients Coef in the classifier;
step 3: generation of a predictive model: and (3) constructing a lymph node metastasis prediction model by adopting the gene mutation characteristics obtained in the step (1) and the mIF image characteristics obtained in the step (2).
Step 4: and (3) verifying a prediction model: and verifying the accuracy, specificity, sensitivity and robustness of the prediction model.
Alternatively, in the above prediction model construction method,
in the step 1, systematically analyzing the gene mutation spectrum of a lung cancer patient, calculating a maximal linear classifier of the interval of lymph node metastasis or not by using a support vector machine (support vector machines, SVM) based on various gene mutation characteristics, thereby obtaining the coefficient Coef of each gene mutation characteristic in the classifier, and obtaining one or more gene mutation characteristics with larger absolute values of Coef;
in step 2.2, shooting detailed images of the stained tissue sections, unmixing the multispectral images by utilizing a spectral library constructed by the monoclonal antibody stained tissue images of each antigen, splicing the images of each field of view to obtain a complete view of the sections, segmenting the images by using software to generate mask files, analyzing by using software, and obtaining the distribution of various single cells on the sections by clustering analysis;
in step 2.3, defining tumor immunity microenvironment characteristics of the patient according to four dimensions of a closest inter-cell nucleus distance (distMin), a farthest inter-cell nucleus distance (distMax), an average inter-cell nucleus distance (distMean) and cell attributes;
in step 2.4, using SVM to calculate the interval maximum linear classifier of lymph node metastasis or not, thereby obtaining the coefficient Coef of each image feature in the classifier, and obtaining one or more mIF image features with larger Coef absolute value;
in step 3, the SVM parameters of GeneFeaturs and ImFeaturs are integrated by using the gene mutation features obtained in step 1 and the mIF image features obtained in step 2, and the optimal parameters are screened by using a Leave-One-Out Cross-Validation method, and then a lymph node metastasis prediction model is constructed by using Logistic regression.
In a more preferred embodiment, the prediction model construction method includes the steps of:
step 1: obtaining the mutation characteristics of the genes:
systematically analyzing the gene mutation spectrum of a lung cancer patient, calculating a maximum interval linear classifier for lymph node metastasis or not by using a support vector machine (support vector machines, SVM) based on various gene mutation characteristics, thereby obtaining coefficients Coef of the various gene mutation characteristics in the classifier, and obtaining one or more gene mutation characteristics with larger absolute values of Coef;
step 2: acquiring mIF image characteristics:
step 2.1: subjecting a lung cancer tissue section of a patient to mIF using corresponding antibodies to one or more lung cancer cell surface antigens;
step 2.2: shooting detailed images of the dyed tissue slices, unmixing the multispectral images by utilizing a spectral library constructed by the monoclonal antibody dyed tissue images of each antigen, splicing the images of each view to obtain a complete view of the slices, segmenting the images by using software to generate mask files, analyzing by using software, and obtaining the distribution of various single cells on the slices by clustering analysis;
step 2.3: defining tumor immunity microenvironment characteristics of patients according to four dimensions of a closest inter-cell distance (distMin), a farthest inter-cell distance (distMax), an average inter-cell distance (distMean) and cell attributes, counting 9 ten digits of all the patients on the characteristics for each dimension of the characteristics, wherein the 9 ten digits divide each dimension of the characteristics into 10 intervals, namely, for the characteristics A, the characteristics A are divided into 10 intervals according to the overall situation of all the patients:if the characteristic A of a certain patient is counted, the proportion of the characteristic A falling in the 10 intervals is counted, and the 1-dimensional characteristic is diffused into 10 dimensions, and is marked as A_0, A_1 and … A_9;
step 2.4: calculating an interval maximum linear classifier of lymph node metastasis or not by using an SVM (support vector machine), so as to obtain coefficients Coef of each image feature in the classifier, and obtaining one or more mIF image features with larger Coef absolute values;
step 3: generation of a predictive model:
and (3) integrating the SVM parameters of GeneFeaturs and ImFeaturs by using the gene mutation characteristics obtained in the step (1) and the mIF image characteristics obtained in the step (2), screening the optimal parameters by using a Leave-One-Out Cross-Validation method, and constructing a lymph node metastasis prediction model by using Logistic regression.
Step 4: and (3) verifying a prediction model: and verifying the accuracy, specificity, sensitivity and robustness of the prediction model.
Alternatively, in the method for constructing a predictive model, the lung cancer is lung cancer stage I-IV.
Alternatively, in the above prediction model construction method, in step 2.1, the lung cancer cell surface antigen used is selected from one or more of the following: PANCK, CD4, CD8, FOXP3, CD68, CD86, MPO, CD11c, CD11b, CD206, PD-1, CD19, CD3, aasma, CD31, collagen I, CD94, or PD-L1.
Preferably, in step 2.1, the lung cancer cell surface antigen used is selected from the group consisting of PANCK, CD4, CD8, FOXP3, CD68 and PD-L1.
Alternatively, in the above prediction model building method, in step 2.2, the resolution of the image is 4028×3012 pixels, the image analysis software used is the form advanced image analysis software, the image is segmented by using CellProfiler software, the mask file is analyzed by using histoCAT software, and the clustering method used is a supervised pedigree attribution method.
Alternatively, in the above prediction model construction method, in step 2.2, the distribution of a plurality of different single cells on the section is clustered by a supervised lineage attribution method, the single cells being selected from two or more of: macrophage & CD8T, macrophage & T, CD T: CD4+FOXP3-PDL1-, CD8T: cd8+pdl1-, macrophages: cd68+pdl1-, other, treg: cd4+ foxp3+, PD-l1+ epithelial cells: panck+pdl1+, epithelial cells: PANCK+PDL1-.
Preferably, the single cell is selected from the group consisting of macrophage & CD8T, macrophage & T, CD4T: CD4+FOXP3-PDL1-, CD8T: cd8+pdl1-, macrophages: cd68+pdl1-, other, treg: cd4+ foxp3+, PD-l1+ epithelial cells: panck+pdl1+ and epithelial cells: PANCK+PDL1-.
Alternatively, in the above prediction model construction method, in step 2.3, the cell property is a cell nucleus area, a long diameter, a short diameter, and a ratio of the long diameter to the short diameter of 4 dimensions, and the intercellular distance parameters in each slice are common (C 2 9 +9) 3=135 dimensions, 139 feature dimensions per slice, 139 x 10=1390 feature dimensions for tumor immune microenvironment based on the mIF slice images for each patient.
Alternatively, in the above prediction model construction method, in step 1, the gene mutation feature includes one or more of the following: KIT, MSH3, APC, HGF, GRIN2A, PARP3, PIK3C2G, KMT2C, TLR4, NOTCH1.
Preferably, the gene mutation is characterized by KIT, MSH3, APC, HGF, GRIN2A, PARP3, PIK3C2G, KMT2C, TLR4, NOTCH1.
Alternatively, in the above prediction model building method, in step 2.4, the meif image feature includes one or more of the following: distmin_treg_treg_4, distmin_ephelial_macromage & cd8t_1, distmin_ephelial_treg_5, distmin_macromage & t_treg_1, distmin_cd8t_macromage_9, distmin_macromage & t_macromage & t_0, distmin_macromage_macromage_3, distmin_macromage & t_macromage & t_2, distmin_cd8t_ephelial_5, distmin_macromage_macromage_4.
Preferably, the mff image features are distmin_treg_treg_4, distmin_ephelial_macromage & cd8t_1, distmin_ephelial_treg_5, distmin_macromage & t_treg_1, distmin_cd8t_macromage_9, distmin_macromage & t_macromage & t_0, distmin_macromage_macromage_3, distmin_macromage & t_macromage & t_2, distmin_cd8t_ephelial_5, distmin_macromage_macromage_4.
In a second aspect, the present invention provides a predictive model system for lung cancer lymph node metastasis constructed by the predictive model construction method described in the first aspect above.
Alternatively, in the above prediction model system, the prediction model system isWherein beta is lm SVM parameters being ImFeaturs, beta Gene Pr (ImGene) is a predicted value of the integrated feature ImGene obtained by integrating the two parameters of GeneFeaturs, and is a probability of lymph node metastasis.
Alternatively, in the predictive model system described above, the genetic mutation signature includes one or more of the following: KIT, MSH3, APC, HGF, GRIN2A, PARP3, PIK3C2G, KMT2C, TLR4, NOTCH1.
Preferably, the gene mutation is characterized by KIT, MSH3, APC, HGF, GRIN2A, PARP3, PIK3C2G, KMT2C, TLR4, NOTCH1.
Alternatively, in the above prediction model system, the mIF image features include one or more of: distmin_treg_treg_4, distmin_ephelial_macromage & cd8t_1, distmin_ephelial_treg_5, distmin_macromage & t_treg_1, distmin_cd8t_macromage_9, distmin_macromage & t_macromage & t_0, distmin_macromage_macromage_3, distmin_macromage & t_macromage & t_2, distmin_cd8t_ephelial_5, distmin_macromage_macromage_4.
Preferably, the mff image features are distmin_treg_treg_4, distmin_ephelial_macromage & cd8t_1, distmin_ephelial_treg_5, distmin_macromage & t_treg_1, distmin_cd8t_macromage_9, distmin_macromage & t_macromage & t_0, distmin_macromage_macromage_3, distmin_macromage & t_macromage & t_2, distmin_cd8t_ephelial_5, distmin_macromage_macromage_4.
Compared with the prior art, the invention has the following beneficial effects:
compared with a single-mode prediction model based on gene mutation characteristics, the multi-mode prediction model system for lung cancer lymph node metastasis constructed by the comprehensive gene mutation characteristics and the mIF image characteristics has higher accuracy, specificity, sensitivity and robustness, and well solves the problems of inaccurate or insufficient sensitivity of the lung cancer lymph node metastasis prediction result in the clinic at present.
Drawings
Other features, objects and advantages of the present disclosure will become more apparent from the following detailed description of non-limiting embodiments, taken in conjunction with the accompanying drawings. In the drawings:
fig. 1: the invention discloses a lung cancer lymph node metastasis multi-mode prediction model schematic diagram integrating gene mutation characteristics and mIF image characteristics.
Fig. 2: in the analysis of gene mutation characteristics and transfer correlation, 10 genes with the maximum absolute value of Coef are obtained.
Fig. 3: schematic of the distribution of 9 different single cells on sections clustered by supervised lineage assignment method.
Fig. 4: representative photographs of the distribution of 9 different single cells on sections clustered by supervised lineage assignment method. Wherein the magnification is 20 times.
Fig. 5: in distMin image feature and transfer correlation analysis, the absolute value of Coef is the largest of the 10 genes.
Fig. 6: a calculation formula schematic diagram of a multi-mode prediction model of lung cancer lymph node metastasis.
Fig. 7: the multimodal predictive model of the present invention predicts ROC curves for lymph node metastasis.
Fig. 8: the multimodal predictive model (validation set) of the present invention predicts ROC curves for lymph node metastasis.
Fig. 9: AUC results for single mode predictive models reported in the literature heretofore.
Detailed Description
The invention will be further illustrated with reference to specific examples. It should be understood that the detailed description and specific examples are intended for purposes of illustration only and are not intended to limit the scope of the invention.
In this disclosure, it is to be understood that terms such as "comprises" or "comprising," etc., are intended to indicate the presence of a tag, number, step, action, component, section, or combination thereof disclosed in this specification, and are not intended to exclude the possibility that one or more other tags, numbers, steps, actions, components, sections, or combinations thereof, are present or added.
The specific techniques or conditions are not identified in the examples and are described in the literature in this field or are carried out in accordance with the product specifications. The reagents or equipment used were conventional products available for purchase through regular channels, with no manufacturer noted.
The experimental methods in the following examples are conventional methods unless otherwise specified. The test materials used in the examples described below, unless otherwise specified, are all commercially available products.
Examples
In detecting the gene mutation signatures (GeneFeatures), gene mutation patterns of tumor tissues were systematically analyzed based on NGS data of 257 lung adenocarcinoma patients (clinical trial data from the university of beijing people hospital, IRB No. 2021PHB182-001), and a maximally spaced linear classifier of lymph node metastasis or not was calculated based on the various gene mutation signatures using a support vector machine (support vector machines, SVM), thereby obtaining coefficients Coef of the respective gene mutation signatures in the classifier (table 1). Where Coef >0 represents the positive correlation between the feature and the transition, coef <0 represents the positive correlation between the feature and the transition, and the absolute value size represents the degree of correlation. The 10 genes with the largest absolute value of Coef included KIT, MSH3, APC, HGF, GRIN2A, PARP3, PIK3C2G, KMT2C, TLR4, NOTCH1, etc. (fig. 2).
Table 1: coefficient of gene mutation characteristics
In terms of tumor immunomicroenvironment, the present invention uses corresponding antibodies to 6 cell surface antigens including PANCK, CD4, CD8, FOXP3, CD68, and PD-L1 to perform multiple immunofluorescent staining (mIF) on lung cancer tissue sections.
For analysis of tissue sections PerkinElmer Vectra (Vectra 3.0.5) was used, the whole section was previewed with a 4 x microscope objective lens and detailed images were taken with a 20 x microscope objective lens. The resolution of the obtained image is up to 4028 multiplied by 3012 pixels, which is beneficial to accurate detection. The multispectral images were then unmixed using the form advanced image analysis software (version 2.3.0) using a spectral library constructed from monoclonal antibody stained tissue images for each antigen. The images of each field of view are then stitched to obtain a complete view of the slice. Finally, we used CellProfiler software to segment the image, generate mask files, and analyzed using histoCAT software, clustered by supervised lineage attribution method to obtain the distribution of 9 different single cells on the sections (see fig. 3 and 4, respectively). As shown in fig. 3, the 9 different single cells were macrophage & CD8T, macrophage & T, CD T: CD4+FOXP3-PDL1-, CD8T: cd8+pdl1-, macrophages: cd68+pdl1-, other, treg: cd4+ foxp3+, PD-l1+ epithelial cells: panck+pdl1+, epithelial cells: PANCK+PDL1-.
To evaluate TIME further based on the image features (image) of the mIF slices, tumor immune microenvironment features for each patient were defined in terms of four dimensions, inter-nuclei closest distance (distMin), inter-nuclei farthest distance (distMax), inter-nuclei mean distance (distMean), and cell attributes. distMin comprises 9 nearest distances between the nuclei of each cell and the nuclei of the same cell, other 8 cells, respectively, such as the nearest distance parameter between macrophages (Macrophage) and the same class of macrophages is denoted distmin_macrophage_macrophage, and the nearest distance between cd4+ T cells (CD 4T) and macrophages is denoted distmin_cd4t_macrophage, distMax and distMean, respectively, so that the intercellular distance parameter in one slice is common (C 2 9 +9)3 = 135 dimensions; while the cell attributes include 4 dimensions of nuclear area, major diameter, minor diameter, and ratio of major to minor diameters. So there are 135+4=139 feature dimensions for each slice.
To fully characterize the heterogeneity between tumors of different patients, for each dimension of the image feature, 9 ten digits of all patients on the image feature are counted, and the 9 ten digits divide each dimension of the image feature into 10 intervals, namely, for the image feature a, the image feature a is divided into 10 intervals according to the overall situation of all patients:. If the image feature A of a certain patient is counted, the proportion of the image feature A falling in the 10 intervals is counted, and the 1-dimensional image feature is diffused into 10 dimensions and is marked as A_0, A_1 and ⋯ A_9. The image feature dimension of the TIME is thus estimated based on the mIF slice for each patient to be expanded to 139 x 10 = 1390 dimensions.
To evaluate the effect of TIME on lymph node metastasis in lung cancer patients, a Support Vector Machine (SVM) was used to calculate a maximal-interval linear classifier of lymph node metastasis or not based on the 1390-dimensional image features reflected by the aforementioned mIF slices, thereby obtaining coefficients Coef (table 2) of the respective image features in the classifier, representing the predicted intensities of various immune cell-cell interactions for lymph node metastasis. The 10 mIF image features with the largest Coef absolute values include distMin_Treg_Treg_4, distMin_Epithelial_macromage & CD8T_1, distMin_Epithelial_Treg_5, distMin_macromage & T_Treg_1, distMin_CD8T_macromage_9, distMin_macromage & T_macromage & T_0, distMin_macromage_macromage_3, distMin_macromage & T_macromage & T_2, distMin_CD8T_Epithelial_5, distMin_macromage_4 (FIG. 5).
Table 2: coefficients of mIF image features
The genetic mutation features and immune microenvironment mIF image features were combined, the SVM parameters of GeneFeaturs and ImFeaturs were integrated, and the best parameters were screened using the Leave-One-Out Cross-Validation method, and then a lymph node metastasis prediction model was constructed using Logistic regression (FIG. 6). Where βim is the SVM parameter of ImFeatures, βgene is the parameter of GeneFeatures, and Pr (ImGene) is the predicted value of the integrated feature ImGene obtained by integrating both of them for the probability of lymph node metastasis.
Specifically, as shown below, to predict LN staging, a library in scikit-learn (1.0.1, https:// scikit-learn. Org/stable /) Python was used. The first 10 coefficients were selected from the mhic image features (ImFeatures) and the genomic features (GeneFeatures) to construct the model (as shown in tables 1 and 2). After removal of imfeatenes mismatched samples, a dataset consisting of 83 mIHC and NGS patients was obtained. Among the classifiers, the ImGene Support Vector Machine (SVM) exhibited the highest performance, AUC of 0.86, F1 score of 0.83, accuracy of 0.82, exceeding Random Forest, imFeatures SVM, geneFeatures SVM and Gradient Boosting.
The ROC curve of the comprehensive characteristics for lymph node metastasis prediction is drawn, and the area under the curve (AUC) is as high as 0.86, which shows that the lymph node metastasis prediction model constructed based on the comprehensive characteristics has good prediction effect (figure 7).
In addition, the above prediction model was also verified. An independent external dataset was collected, comprising 61 patients receiving complete excision therapy, ethical No.2021PHB182-001. All patients provided informed consent for sample collection and all participants agreed to publish study results. These patients were selected based on the availability of NGS and mIHC data. The AUC for the independent external data set was 0.81 and the training queue 95% confidence intervals were AUC [0.624,1.036] and [0.457,0.963] (fig. 8).
Therefore, the multi-modal prediction model of the lung cancer lymph node metastasis has better prediction effect on lymph node metastasis than the single-modal prediction model reported in the previous literature. Existing lymph node metastasis prediction models are mainly single-mode prediction models, such as PKU models (Chen K, yang F, jiang G, li J, wang J Development and validation of a clinical prediction model for N2 lymph node metastasis in non-small cell lung cancer, ann Thorac Surg.2013 nov;96 (5): 1761-8), FUDAN models (Zhang Y, sun Y, xang J, zhang Y, hu H, chen H. A prediction model for N2 disease in T1 non-small cell lung cancer, J Thorac Cardiovasc Surg.2012 Dec;144 (6): 1360-4) based solely on clinical characteristics of the patient with AUC of 0.767, 0.806, respectively, sensitivity and specificity being low. The lung cancer multi-mode lymph node metastasis prediction model constructed by combining the gene mutation characteristics and the mIF image characteristics has AUC as high as 0.86, and has higher effect than the specificity and sensitivity of the existing prediction model (figure 9).
Compared with a single-mode prediction model based on gene mutation characteristics, the multi-mode prediction model system for lung cancer lymph node metastasis constructed by the comprehensive gene mutation characteristics and the mIF image characteristics has higher accuracy, specificity, sensitivity and robustness, and well solves the problems of inaccurate or insufficient sensitivity of the lung cancer lymph node metastasis prediction result in the clinic at present.
The foregoing description is only of the preferred embodiments of the present disclosure and description of the principles of the technology being employed. It will be appreciated by those skilled in the art that the scope of the invention referred to in this disclosure is not limited to the specific combination of features described above, but encompasses other embodiments in which any combination of features described above or their equivalents is contemplated without departing from the inventive concepts described. Such as those described above, are provided in the present disclosure in place of, but not limited to, features having similar functions.
Claims (8)
1. A method for constructing a multi-modal prediction model for lung cancer lymph node metastasis by combining gene mutation characteristics and multiple immunofluorescence staining mIF image characteristics is characterized by comprising the following steps: the prediction model construction method comprises the following steps:
step 1: obtaining the mutation characteristics of the genes: systematically analyzing the gene mutation spectrum of a lung cancer patient, calculating a maximum interval linear classifier for lymph node metastasis or not by using a support vector machine based on various gene mutation characteristics, thereby obtaining coefficients Coef of the various gene mutation characteristics in the classifier, and obtaining one or more gene mutation characteristics with larger absolute values of Coef;
step 2: acquiring mIF image characteristics:
step 2.1: subjecting a lung cancer tissue section of a patient to mIF using corresponding antibodies to one or more lung cancer cell surface antigens;
step 2.2: shooting detailed images of the dyed tissue slices, unmixing the multispectral images by utilizing a spectral library constructed by the monoclonal antibody dyed tissue images of each antigen, splicing the images of each view to obtain a complete view of the slices, segmenting the images by using software to generate mask files, analyzing by using software, and obtaining the distribution of various single cells on the slices by clustering analysis;
step 2.3: evaluating the influence of the tumor immune microenvironment on the lymph node metastasis of a lung cancer patient, and defining the tumor immune microenvironment characteristics of the patient according to four dimensions, namely the closest distance between cell nuclei, the farthest distance between cell nuclei, the average distance between cell nuclei and cell attributes;
step 2.4: obtaining one or more mIF image features with larger absolute values of Coef coefficients in the classifier, calculating an interval maximum linear classifier of lymph node metastasis or not by using an SVM, thereby obtaining coefficients Coef of each image feature in the classifier, and obtaining one or more mIF image features with larger absolute values of Coef;
step 3: generation of a predictive model: integrating SVM parameters of GeneFeaturs and ImFeaturs by adopting the gene mutation characteristics obtained in the step 1 and the mIF image characteristics obtained in the step 2, screening optimal parameters by using a Leave-One-Out Cross-Validation method, and constructing a lymph node metastasis prediction model by using Logistic regression;
step 4: and (3) verifying a prediction model: verifying the accuracy, specificity, sensitivity and robustness of the predictive model,
the prediction model isWherein beta is lm SVM parameters being ImFeaturs, beta Gene Pr (ImGene) is a predicted value of the integrated feature ImGene obtained by integrating the two parameters of GeneFeaturs, and is a probability of lymph node metastasis.
2. The prediction model construction method according to claim 1, characterized in that: the lung cancer is lung cancer stage I-IV.
3. The prediction model construction method according to claim 1, characterized in that: in step 2.1, the lung cancer cell surface antigen used is selected from one or more of the following: PANCK, CD4, CD8, FOXP3, CD68, CD86, MPO, CD11c, CD11b, CD206, PD-1, CD19, CD3, aasma, CD31, collagen I, CD94, or PD-L1.
4. The prediction model construction method according to claim 1, characterized in that: in step 2.2, the distribution of a plurality of different single cells on the sections is obtained by supervised lineage homing method clustering, the single cells being selected from two or more of the following: macrophage & CD8T, macrophage & T, CD T: CD4+FOXP3-PDL1-, CD8T: cd8+pdl1-, macrophages: cd68+pdl1-, other, treg: cd4+ foxp3+, PD-l1+ epithelial cells: panck+pdl1+, epithelial cells: PANCK+PDL1-.
5. The prediction model construction method according to claim 1, characterized in that: in step 2.3, the cell properties are the cell nucleus area, the long diameter, the short diameter and the ratio of the long diameter to the short diameter of 4 dimensions, and the intercellular distance parameters in each slice are common (C 2 9 +9) 3=135 dimensions, 139 feature dimensions per slice, 139 x 10=1390 feature dimensions for tumor immune microenvironment based on the mIF slice images for each patient.
6. The prediction model construction method according to claim 1, characterized in that: in step 1, the genetic mutation signature comprises one or more of the following: KIT, MSH3, APC, HGF, GRIN2A, PARP3, PIK3C2G, KMT2C, TLR4, NOTCH1.
7. The prediction model construction method according to claim 1, characterized in that: in step 2.4, the mIF image features include one or more of: distmin_treg_treg_4, distmin_ephelial_macromage & cd8t_1, distmin_ephelial_treg_5, distmin_macromage & t_treg_1, distmin_cd8t_macromage_9, distmin_macromage & t_macromage & t_0, distmin_macromage_macromage_3, distmin_macromage & t_macromage & t_2, distmin_cd8t_ephelial_5, distmin_macromage_macromage_4.
8. A predictive model system for lung cancer lymph node metastasis constructed by the predictive model construction method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311524786.2A CN117275578B (en) | 2023-11-16 | 2023-11-16 | Method for constructing multi-mode prediction model of lung cancer lymph node metastasis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311524786.2A CN117275578B (en) | 2023-11-16 | 2023-11-16 | Method for constructing multi-mode prediction model of lung cancer lymph node metastasis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117275578A CN117275578A (en) | 2023-12-22 |
| CN117275578B true CN117275578B (en) | 2024-02-27 |
Family
ID=89214581
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311524786.2A Active CN117275578B (en) | 2023-11-16 | 2023-11-16 | Method for constructing multi-mode prediction model of lung cancer lymph node metastasis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117275578B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111521794A (en) * | 2020-04-21 | 2020-08-11 | 山东第一医科大学(山东省医学科学院) | Immunofluorescence kit and detection method for detecting NSE gene mutation of peripheral blood circulating tumor cells of small cell lung cancer patients |
| WO2021174052A1 (en) * | 2020-02-27 | 2021-09-02 | Foundation Medicine, Inc. | Mitigation of statistical bias in genetic sampling |
| CN115424740A (en) * | 2022-09-30 | 2022-12-02 | 四川大学华西医院 | Tumor immunotherapy effect prediction system based on NGS and deep learning |
| CN116047074A (en) * | 2022-11-08 | 2023-05-02 | 兰州大学第一医院 | Marker for diagnosing and/or predicting lung cancer, diagnostic model and construction method thereof |
| CN116516004A (en) * | 2023-03-15 | 2023-08-01 | 中南大学湘雅二医院 | Predictive model for driving malignant lung nodule lymph node metastasis by tumor genes through changing immune microenvironment |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20200199555A1 (en) * | 2018-12-05 | 2020-06-25 | The Broad Institute, Inc. | Cas proteins with reduced immunogenicity and methods of screening thereof |
-
2023
- 2023-11-16 CN CN202311524786.2A patent/CN117275578B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021174052A1 (en) * | 2020-02-27 | 2021-09-02 | Foundation Medicine, Inc. | Mitigation of statistical bias in genetic sampling |
| CN111521794A (en) * | 2020-04-21 | 2020-08-11 | 山东第一医科大学(山东省医学科学院) | Immunofluorescence kit and detection method for detecting NSE gene mutation of peripheral blood circulating tumor cells of small cell lung cancer patients |
| CN115424740A (en) * | 2022-09-30 | 2022-12-02 | 四川大学华西医院 | Tumor immunotherapy effect prediction system based on NGS and deep learning |
| CN116047074A (en) * | 2022-11-08 | 2023-05-02 | 兰州大学第一医院 | Marker for diagnosing and/or predicting lung cancer, diagnostic model and construction method thereof |
| CN116516004A (en) * | 2023-03-15 | 2023-08-01 | 中南大学湘雅二医院 | Predictive model for driving malignant lung nodule lymph node metastasis by tumor genes through changing immune microenvironment |
Non-Patent Citations (1)
| Title |
|---|
| Wei W ,Li X ,Song M , et al.Molecular Analysis of Oncogenic Mutations in Resected Margins by Next-Generation Sequencing Predicts Relapse in Non-Small Cell Lung Cancer Patients.OncoTargets and therapy.2020,(第2020,13期),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117275578A (en) | 2023-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112185569B (en) | Breast cancer patient disease-free survival period prediction model and construction method thereof | |
| Maldonado et al. | Noninvasive characterization of the histopathologic features of pulmonary nodules of the lung adenocarcinoma spectrum using computer-aided nodule assessment and risk yield (CANARY)—a pilot study | |
| Xu et al. | Using transfer learning on whole slide images to predict tumor mutational burden in bladder cancer patients | |
| Baik et al. | Automated classification of oral premalignant lesions using image cytometry and random forests-based algorithms | |
| US20140235487A1 (en) | Oral cancer risk scoring | |
| Siviengphanom et al. | Mammography-based radiomics in breast cancer: a scoping review of current knowledge and future needs | |
| CN116188423A (en) | Super-pixel sparse and unmixed detection method based on pathological section hyperspectral image | |
| Korkmaz et al. | Least square support vector machine and minumum redundacy maximum relavance for diagnosis of breast cancer from breast microscopic images | |
| CN114066882A (en) | Lung adenocarcinoma Ki67 expression level non-invasive detection method and device based on depth imaging omics | |
| CN111833963A (en) | cfDNA classification method, device and application | |
| Beig et al. | Radiogenomic analysis of hypoxia pathway reveals computerized MRI descriptors predictive of overall survival in glioblastoma | |
| Zhu et al. | A computerized tomography-based radiomic model for assessing the invasiveness of lung adenocarcinoma manifesting as ground-glass opacity nodules | |
| Smedley et al. | Using deep neural networks and interpretability methods to identify gene expression patterns that predict radiomic features and histology in non-small cell lung cancer | |
| Pan et al. | Application of radiomics in diagnosis and treatment of lung cancer | |
| Yang et al. | Deep learning model for the detection of prostate cancer and classification of clinically significant disease using multiparametric MRI in comparison to PI-RADs score | |
| CN117275578B (en) | Method for constructing multi-mode prediction model of lung cancer lymph node metastasis | |
| Huang et al. | Coupling radiomics analysis of CT image with diversification of tumor ecosystem: A new insight to overall survival in stage I− III colorectal cancer | |
| CN116504406A (en) | Method and system for constructing lung cancer postoperative risk model based on image combination pathology | |
| CN110942808A (en) | Prognosis prediction method and prediction system based on gene big data | |
| Emaminejad et al. | Applying a radiomics approach to predict prognosis of lung cancer patients | |
| Zargar et al. | Using VGG16 Algorithms for classification of lung cancer in CT scans Image | |
| CN117275744B (en) | Method for constructing lung cancer prognosis multi-mode prediction model by combining gene mutation characteristics and mIF image characteristics | |
| US20130080101A1 (en) | System, method and computer-accessible medium for evaluating a malignancy status in at-risk populations and during patient treatment management | |
| Pham et al. | Deep learning Of P73 biomarker expression in rectal cancer patients | |
| Huang et al. | Conditional generative adversarial network driven radiomic prediction of mutation status based on magnetic resonance imaging of breast cancer |
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 | ||
| CB03 | Change of inventor or designer information |
Inventor after: Li Hao Inventor after: Chen Kezhong Inventor after: Wang Jun Inventor after: Yang Fan Inventor after: Li Yun Inventor after: Sheng Jianpeng Inventor before: Li Hao Inventor before: Wang Jun Inventor before: Yang Fan Inventor before: Li Yun Inventor before: Sheng Jianpeng |
|
| CB03 | Change of inventor or designer information | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240724 Address after: Room 301, 401, Building 11, Qianwan Bioport Phase I, Life Science and Technology Innovation Center, No. 3300 Benjing Avenue, Ningwei Street, Xiaoshan District, Hangzhou City, Zhejiang Province 311200 Patentee after: Hangzhou Yinfei Duozuo Biotechnology Co.,Ltd. Country or region after: China Address before: 100044 No. 11 South Main Street, Xicheng District, Beijing, Xizhimen Patentee before: PEKING UNIVERSITY PEOPLE'S Hospital Country or region before: China |
|
| TR01 | Transfer of patent right |