CN115527683A - Method for predicting targeted therapeutic effect of Lunvatinib of liver cancer patient based on artificial intelligence - Google Patents
Method for predicting targeted therapeutic effect of Lunvatinib of liver cancer patient based on artificial intelligence Download PDFInfo
- Publication number
- CN115527683A CN115527683A CN202211222615.XA CN202211222615A CN115527683A CN 115527683 A CN115527683 A CN 115527683A CN 202211222615 A CN202211222615 A CN 202211222615A CN 115527683 A CN115527683 A CN 115527683A
- Authority
- CN
- China
- Prior art keywords
- artificial intelligence
- liver cancer
- tumor
- patients
- lunvatinib
- 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.)
- Withdrawn
Links
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- 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]
-
- 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
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- 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
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
-
- 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
- G06V10/54—Extraction of image or video features relating to texture
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/40—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for data related to laboratory analysis, e.g. patient specimen analysis
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H30/00—ICT specially adapted for the handling or processing of medical images
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H80/00—ICT specially adapted for facilitating communication between medical practitioners or patients, e.g. for collaborative diagnosis, therapy or health monitoring
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10081—Computed x-ray tomography [CT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30056—Liver; Hepatic
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30096—Tumor; Lesion
-
- 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)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Theoretical Computer Science (AREA)
- Public Health (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Software Systems (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Artificial Intelligence (AREA)
- Radiology & Medical Imaging (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Databases & Information Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method for predicting the targeted therapeutic effect of Lunvatinib of a liver cancer patient based on artificial intelligence, which comprises the following steps: (1) collecting data of patients with advanced liver cancer; (2) Judging the disease progression and the drug treatment effect of patients, grouping the patients according to the evaluation criteria of solid tumors, counting and analyzing two groups of differences; (3) Manually segmenting the patient imaging picture by using image processing software to obtain a tumor 3D model; (4) Extracting and quantifying the characteristics of the tumor image by Python software; (5) Screening image features by using an LASSO regression model, eliminating highly correlated features by using a correlation matrix, and leaving a non-redundant feature set; (6) And analyzing the extracted imaging characteristics and clinical data by using an artificial intelligence machine learning method, and establishing a prediction model based on the imaging characteristics. The method utilizes an artificial intelligence technology to predict the sensitivity of the targeted therapy of the Rankine, and has important clinical application value.
Description
Technical Field
The invention relates to a method for predicting the targeted therapeutic effect of Lunvatinib of a late-stage liver cancer patient based on artificial intelligence image omics.
Background
Primary liver cancer is one of the most common types of cancer in the world. Hepatocellular carcinoma (HCC) accounts for about 75-85% of primary liver cancer, and risk factors thereof include hepatitis B, alcohol abuse, smoking, liver cirrhosis, and the like. The prognosis of liver cancer patients is very poor, the initial surgical resection rate is only 15% -30%, and most patients cannot be subjected to radical surgical resection after diagnosis. For the liver cell liver cancer patients who can not be removed by operation, the treatment method comprises the traditional methods of chemotherapy, interventional therapy, local therapy, and the methods of targeted therapy, immunotherapy and the like which are started in recent years, although the overall survival rate of the patients is improved, the overall effective rate of the system therapy is only less than 30%, most liver cancer patients who can not be removed by operation can not benefit from the system therapy, and the survival prognosis of the patients is seriously influenced. Lovatinib, one of targeted therapeutic drugs for patients with hepatocellular carcinoma, is a multi-kinase inhibitor with VEGFR1-3, FGFR1-4, PDGFR-alpha, c-Kit and RET as targets, and is classified as a first-line therapeutic Drug for patients with advanced hepatocellular carcinoma by the Food and Drug Administration (FDA) and the national anticancer society's clinical oncology collaborative professional committee (CSCO). In past global multicenter phase iii clinical trials, rivastigmine demonstrated non-inferiority and higher objective remission rates than the traditional classical targeted drug sorafenib, and is currently more selected for use by clinically moderate patients with advanced liver cancer because of its smaller toxic side effects and longer survival benefit. However, because of the high heterogeneity of liver cancer and the individual differences among different patients, the overall effective rate of the ranvatinib medicament is still unsatisfactory, the overall effective rate is only about 2% -18.8%, most of patients taking the ranvatinib cannot be objectively relieved, and the median survival time of the liver cancer patients is only 10.7-11.8 months. Therefore, how to predict the sensitivity of the liver cell liver cancer patient to the Rankine medicament in early stage is expected to adjust the treatment scheme in early stage, so that the liver cancer patient with later stage can be guided to take more accurate medicament, the overall treatment effect of the medicament and the survival prognosis of the patient are improved, the method is a problem which needs to be solved urgently in clinic, and has very important clinical practice significance.
In recent 10 years, the development of artificial intelligence technology and image processing technology has been rapid, which promotes the generation of image omics technology and gradually becomes an important component of artificial intelligence applied in the medical diagnosis and treatment field. Image recognition is an important field of artificial intelligence, and refers to a technology for processing, analyzing and understanding images by using a computer to recognize various targets and objects in different modes. The technology of imaging omics (radiomics) derived from the image recognition technology is increasingly gaining attention, and the imaging omics are used for extracting and quantifying massive characteristic data from medical images (images such as ultrasound, X-ray, CT, MRI or PET and the like) by using an automatic data characteristic extraction algorithm, analyzing different clinical phenotypes such as patient genotyping, treatment curative effect, clinical results and the like by using the same, and finally guiding clinical decision to realize precise medicine. Since 2016, the imaging technology has been widely used in the fields of tumor detection, diagnosis, pathological grading, tumor efficacy evaluation, and tumor prognosis, and has shown high clinical value and clinical acceptance.
In recent years, the rapid advancement of biomedical technologies, particularly gene sequencing technologies, has led to the continuous identification of features in tumors based on molecular changes. With the deep understanding of the biological behavior of tumors, the heterogeneity of tumors becomes an important barrier to the therapeutic effect of anti-tumors, which seriously affects the therapeutic effect and the prognosis of patients. Since the introduction of the concept of precision medicine, various branches of the medical field have entered the precision medicine era, particularly in oncology research. Precision medicine is widely used in cancer research, and research results are also transformed and used for clinical tumor treatment. In the precise medical age, how to achieve the 'individuation, predictability, precision, prevention and participation' principle of tumor diagnosis and treatment, improve the curative effect of treatment and improve the survival prognosis of patients becomes a hotspot and difficulty in the field of current medical research.
At present, no relevant research is available in the medical field, the research on the aspect of predicting the accurate treatment of the late-stage liver cancer patient by the artificial intelligence imaging omics technology is explored, and the problem to be solved by the invention is how to realize the prediction of the targeted treatment sensitivity of the Lunatinib of the late-stage liver cancer patient and guide the medication scheme of the clinical patient by the artificial intelligence imaging omics technology.
Disclosure of Invention
Aiming at the defects and the study blind areas of the prior art, the invention provides a method for predicting the target treatment effect of the ranvatinib for the liver cancer patient based on artificial intelligence.
In order to achieve the aim, the invention elaborates a method for predicting the target treatment effect of the Lunvatinib of the liver cancer patient based on artificial intelligence, which comprises the following steps:
(1) Collecting imaging data and clinical data of patients with advanced liver cancer taking Lunvatinib in a plurality of medical centers, wherein the imaging data comprise image pictures of arterial phase and venous phase enhanced CT, the clinical data comprise epidemiological data, laboratory examination data and tumor related data, the epidemiological data comprise sex, age, height, weight, BMI, diabetes, hypertension and PS scores, the laboratory examination data comprise hepatitis B, liver cirrhosis, liver function indexes, tumor indexes and blood coagulation function indexes, and the tumor related data comprise tumor size, tumor number, tumor stage, lymph node metastasis and vascular invasion;
(2) Performing outpatient follow-up visits on patients once a month, performing imaging and blood index detection, judging the disease progress and drug treatment effects of the patients, judging the drug treatment effects according to RECIST 1.1 and mRECIST which are solid tumor evaluation standards, dividing the patients into objective remission groups and non-objective remission groups according to the drug treatment effects, counting clinical data between the two groups, and performing single-factor and multi-factor logistic regression analysis;
(3) Image segmentation: inviting 2-3 imaging physicians with professional experience for more than 10 years, manually segmenting the enhanced CT image pictures of the arterial phase and the venous phase of the patient by using MRIcroGL software, and obtaining a 3D model of the region of interest;
(4) Image feature extraction and quantization: extracting and quantifying image features of the manually segmented tumor 3D model by Python software, wherein the image features comprise shape features, first histogram features, second histograms and texture features;
(5) Selecting characteristics: screening image features by using an LASSO regression model, eliminating highly correlated features by using a correlation matrix, and remaining a non-redundant feature set for subsequent analysis and modeling;
(6) And analyzing the extracted imaging characteristics and clinical data by using an artificial intelligence machine learning method, and establishing a prediction model based on the imaging characteristics for predicting the effectiveness of the Ranuncutinib.
Further, the step (1) further comprises: and screening the patients by adopting inclusion criteria and exclusion criteria, and selecting the patients meeting the inclusion criteria for subsequent analysis so as to exclude the influence of other interference factors on the prediction result.
Further, in the classification criteria in said step (2), the tumor response is defined as complete remission, partial remission, disease stabilization and disease progression, the objective remission group is defined as patients in complete remission and partial remission, and the objective remission free group is defined as patients in disease stabilization and disease progression.
Further, in the step (3), consistency check is performed on the image segmentation models among different physicians, so as to ensure the stability and repeatability of the extracted imaging features, and if there is any discrepancy, consensus is achieved through discussion.
Further, in the step (4), the extracted features include tumor morphology, tumor margin, intensity, haralik, invariant moment, and discrete wavelet transform features, and the extracted features are normalized by a Z-score method.
Further, in the step (6), the artificial intelligence machine learning method includes logistic regression, decision tree, random forest, neural network, support vector machine, bootstrap method, cross validation, and cluster analysis method.
Further, the step (6) further comprises: the reliability and repeatability of the result are ensured by using a 10-times cross validation machine learning method to avoid the over-fitting phenomenon.
Further, the step (6) further comprises: comprehensively evaluating the prediction effectiveness of the established model through a plurality of evaluation indexes, wherein the evaluation indexes comprise: area under the subject characteristic curve, accuracy, sensitivity, specificity, positive predictive value, and negative predictive value.
Further, the step (6) further comprises: the research objects are divided into training queues and external verification queues, and the effectiveness and the accuracy are further verified in the verification queues through models established by the training queues.
The beneficial effects of the invention are: at present, the research for predicting the effectiveness of the Lunatinib of the late-stage liver cancer patient by using an artificial intelligence imaging omics technology similar to the invention does not exist in the prior art, and the prediction model is established by using an artificial intelligence machine learning method based on the imaging omics characteristics and the clinical characteristics, so that the sensitivity of the Lunatinib of the late-stage liver cancer patient is predicted at an early stage, the selection of a medication strategy of the patient is guided, and the clinical application value is very high. The technology of the invention uses CT imaging data and clinical data which are most commonly used for examination by clinical patients to carry out comprehensive analysis, is a non-invasive, convenient and feasible technical means, has low technical cost and good practicability, and is suitable for medical units of all grades. Meanwhile, the invention uses an artificial intelligence machine learning algorithm to construct a prediction model, thereby avoiding human errors and having higher accuracy and reliability compared with the traditional analysis means. In conclusion, the technology can provide important reference for selecting an accurate treatment scheme of a patient with advanced liver cancer clinically and guide the patient to take medicine, so that the overall survival prognosis of the patient with advanced liver cancer can be improved.
Drawings
FIG. 1 is a design flow diagram of an embodiment of the present invention;
FIG. 2 is a schematic representation of a venous phase enhanced CT image and ROI image segmentation of representative patients with effective and ineffective Lunvatinib treatment for liver cancer;
FIG. 3 is a venous phase enhanced CT image of representative Lunvatinib treatment effective and ineffective liver cancer patients before and after treatment;
figure 4 is a thermographic and clinical baseline profile between the liver cancer patients in the efficacious and ineffective groups of the ranvatinib treatment;
FIG. 5 is a graph of unsupervised clustering machine learning method to predict susceptibility and survival prognosis of Limcatinib-targeted therapy in liver cancer patients;
FIG. 6 is an image omics prediction model ROC curve established based on different machine learning algorithms;
figure 7 is a graph of survival for patients with liver cancer who have been treated with targeted regiment of ranvatinib and not effective.
Detailed Description
The embodiment of the method for predicting the target treatment effect of the Lunvatinib for the liver cancer patient based on artificial intelligence is shown in figures 1-7, and comprises the following steps:
(1) Collecting patients with advanced liver cancer taking ranvatinib in a plurality of medical centers, and screening the patients according to strict inclusion and exclusion criteria, wherein the inclusion criteria comprise: a. a clinical diagnosis of HCC patients; b. no other anti-tumor therapy was received than the targeted therapy with varlitinib; c. the liver function grading child-pugh is less than or equal to 7; d. shooting an enhanced CT image before treatment; e. complete clinical data and follow-up information. Strict inclusion criteria are intended to exclude interference with the results of the study by other clinical factors, thereby improving the confidence and accuracy of the results. And collecting imaging data and clinical data of the patient, wherein the imaging data comprises image pictures of the arterial phase and venous phase enhanced CT before treatment. The clinical data comprise epidemiological data (comprising sex, age, height, weight, BMI, diabetes, hypertension, PS score and the like), laboratory examination data (comprising hepatitis B, cirrhosis, liver function indexes, tumor indexes, blood coagulation function indexes and the like) and tumor-related data (comprising tumor size, tumor number, tumor stage, lymph node metastasis, blood vessel invasion and the like).
(2) The patient is followed up every month, imaging and blood index detection are carried out, the disease progression and the drug treatment effect of the patient are judged, the drug treatment effect is according to the latest solid tumor evaluation criteria RECIST 1.1 and mRECIST, the tumor response is defined as Complete Remission (CR), partial Remission (PR), stable Disease (SD) and disease Progression (PD), the patient can be divided into objective remission group (CR + PR) and non-objective remission group (SD + PD) according to the treatment effect, the clinical data between the two groups are counted, and single-factor and multi-factor logistic regression analysis and survival analysis are carried out.
(3) Image segmentation: inviting 2-3 imaging physicians with professional experience for more than 10 years, manually segmenting the enhanced CT image pictures of the arterial phase and the venous phase of the patient obtained in the steps by using MRIcroGL software (www.mccausslandcenter.sc.edu), obtaining a three-dimensional (3D) structure of a region of interest (ROI), and simultaneously carrying out consistency check on delineation among different physicians so as to measure and calculate the stability and repeatability of the characteristics of the image group.
(4) Image feature extraction and quantization: and extracting and quantifying the shape feature, the first histogram feature, the second histogram feature, the texture feature and the like of the manually segmented tumor 3D model by using Python software.
(5) Selecting characteristics: screening image features by using an LASSO (least absolute shrinkage and selection operator) Cox regression model, eliminating highly relevant features by using a correlation matrix, and remaining a non-redundant feature set for subsequent analysis and modeling;
(6) And (3) analyzing the extracted imaging characteristics and clinical data by using an artificial intelligence machine learning method (such as logistic regression, decision tree, random forest, neural network, support vector machine, bootstrap method, cross validation, cluster analysis method and the like), and establishing a prediction model based on the imaging characteristics for predicting effectiveness of the Ranuncutinib. To better demonstrate the predictive efficacy of the model, a variety of machine learning indicators were used for evaluation, including the area under the subject characteristic curve (AUC), accuracy (ACC), sensitivity, specificity, positive Predictive Value (PPV), and Negative Predictive Value (NPV). In order to verify the reliability and applicability of the model, the study objects are divided into a training queue and an external verification queue, and the model established by the training queue is further verified in the verification queue for validity and accuracy.
The invention has the advantages that:
1. the artificial intelligence technology and the imaging technology are applied to the accurate prediction of the target treatment effect of the Lunvatinib of the late-stage liver cancer patient, and the innovation of the research content is high;
2. in actual application, the system can be researched in a national multi-center queue, and is explored and verified in a plurality of medical centers, so that the data source is real and accurate, and the evidence level is high;
3. the technical content of the invention adopts the imaging CT image and clinical data which are most commonly used for examination of clinical liver cancer patients, and the imaging CT image and the clinical data can be obtained by medical units with different grades, so that the research cost is low and the practicability is good;
3. the invention is used for analyzing and constructing the prediction model by the artificial intelligent machine learning algorithm, avoids the defects of the traditional analysis method and the artificial analysis bias error, and has high accuracy;
4. the method is found through retrieval, the accurate treatment effect of the late-stage liver cancer patient is predicted through the artificial intelligence image omics model for the first time, and the method has very good application value;
5. the invention can provide reference for the accurate treatment of the late-stage liver cancer patient, thereby improving the overall survival prognosis of the late-stage liver cancer patient and having great clinical significance.
The above embodiment is only one of the preferred embodiments of the present invention, and general changes and substitutions by those skilled in the art within the technical scope of the present invention are included in the protection scope of the present invention.
Claims (9)
1. A method for predicting the targeted therapeutic effect of Lunvatinib of a liver cancer patient based on artificial intelligence comprises the following steps:
(1) Collecting imaging data and clinical data of patients with advanced liver cancer taking Lunvatinib in a plurality of medical centers, wherein the imaging data comprise image pictures of arterial phase and venous phase enhanced CT, the clinical data comprise epidemiological data, laboratory examination data and tumor related data, the epidemiological data comprise sex, age, height, weight, BMI, diabetes, hypertension and PS scores, the laboratory examination data comprise hepatitis B, liver cirrhosis, liver function indexes, tumor indexes and blood coagulation function indexes, and the tumor related data comprise tumor size, tumor number, tumor stage, lymph node metastasis and vascular invasion;
(2) Performing outpatient follow-up visits on patients once a month, performing imaging and blood index detection, judging the disease progress and drug treatment effects of the patients, judging the drug treatment effects according to RECIST 1.1 and mRECIST which are solid tumor evaluation standards, dividing the patients into objective remission groups and non-objective remission groups according to the drug treatment effects, counting clinical data between the two groups, and performing single-factor and multi-factor logistic regression analysis;
(3) Image segmentation: inviting 2-3 image physicians with professional experience for more than 10 years, manually segmenting the enhanced CT image pictures of the artery phase and the vein phase of the patient by using MRIcroGL software, and obtaining a 3D model of the region of interest;
(4) Image feature extraction and quantization: extracting and quantifying image features of the manually segmented tumor 3D model by Python software, wherein the image features comprise shape features, first histogram features, second histograms and texture features;
(5) Selecting characteristics: screening image features by using an LASSO regression model, eliminating highly correlated features by using a correlation matrix, and remaining a non-redundant feature set for subsequent analysis and modeling;
(6) And analyzing the extracted imaging characteristics and clinical data by using an artificial intelligence machine learning method, and establishing a prediction model based on the imaging characteristics for predicting the effectiveness of the Ranuncutinib.
2. The method for predicting the effect of targeted therapeutic on Lunvatinib in a patient with liver cancer based on artificial intelligence as claimed in claim 1, wherein: the step (1) further comprises: and screening the patients by adopting inclusion criteria and exclusion criteria, and selecting the patients meeting the inclusion criteria for subsequent analysis so as to exclude the influence of other interference factors on the prediction result.
3. The method for predicting the effect of targeted therapeutic on Lunvatinib in a patient with liver cancer based on artificial intelligence as claimed in claim 1, wherein: in the classification criteria in said step (2), tumor response is defined as complete remission, partial remission, disease stabilization and disease progression, objective remission group is defined as patients in complete remission and partial remission, and objective remission free group is defined as patients in disease stabilization and disease progression.
4. The method for predicting the effect of targeted therapeutic on Lunvatinib in a patient with liver cancer based on artificial intelligence as claimed in claim 1, wherein: in the step (3), consistency check is performed on the image segmentation models among different physicians to ensure the stability and repeatability of the extracted imaging features, and if the extracted imaging features are divergent, consensus is achieved through discussion.
5. The method for predicting the effect of targeted therapeutic on Lunvatinib in a patient with liver cancer based on artificial intelligence as claimed in claim 1, wherein: in the step (4), the extracted features comprise tumor morphology, tumor edge, intensity, haralik, invariant moment and discrete wavelet transform features, and the extracted features are subjected to standardization processing by adopting a Z-score method.
6. The method for predicting the effect of targeted therapeutic on Lunvatinib in a patient with liver cancer based on artificial intelligence as claimed in claim 1, wherein: in the step (6), the artificial intelligence machine learning method comprises logistic regression, decision trees, random forests, neural networks, support vector machines, bootstrap methods, cross validation and cluster analysis methods.
7. The method for predicting the effect of targeted therapy of Lunvatinib on a patient with liver cancer based on artificial intelligence of claim 1, wherein: the step (6) further comprises: the reliability and repeatability of the result are ensured by using a 10-times cross validation machine learning method to avoid the over-fitting phenomenon.
8. The method for predicting the effect of targeted therapy of Lunvatinib on a patient with liver cancer based on artificial intelligence of claim 1, wherein: the step (6) further comprises: comprehensively evaluating the prediction effectiveness of the established model through a plurality of evaluation indexes, wherein the evaluation indexes comprise: area under the subject characteristic curve, accuracy, sensitivity, specificity, positive predictive value, and negative predictive value.
9. The method for predicting the effect of targeted therapy of Lunvatinib on a patient with liver cancer based on artificial intelligence of claim 1, wherein: the step (6) further comprises: the research objects are divided into training queues and external verification queues, and the effectiveness and the accuracy are further verified in the verification queues through models established by the training queues.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211222615.XA CN115527683A (en) | 2022-10-08 | 2022-10-08 | Method for predicting targeted therapeutic effect of Lunvatinib of liver cancer patient based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211222615.XA CN115527683A (en) | 2022-10-08 | 2022-10-08 | Method for predicting targeted therapeutic effect of Lunvatinib of liver cancer patient based on artificial intelligence |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115527683A true CN115527683A (en) | 2022-12-27 |
Family
ID=84701790
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211222615.XA Withdrawn CN115527683A (en) | 2022-10-08 | 2022-10-08 | Method for predicting targeted therapeutic effect of Lunvatinib of liver cancer patient based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115527683A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116758062A (en) * | 2023-08-11 | 2023-09-15 | 之江实验室 | Drug effectiveness evaluation method and device |
| CN117352046A (en) * | 2023-10-13 | 2024-01-05 | 安徽翼方健数信息科技有限公司 | Tumor targeting drug delivery treatment positioning needle system and assembly thereof |
| CN117457188A (en) * | 2023-11-08 | 2024-01-26 | 复旦大学附属中山医院 | System for evaluating histopathological changes in hepatocellular carcinoma caused by non-operative treatment |
-
2022
- 2022-10-08 CN CN202211222615.XA patent/CN115527683A/en not_active Withdrawn
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116758062A (en) * | 2023-08-11 | 2023-09-15 | 之江实验室 | Drug effectiveness evaluation method and device |
| CN117352046A (en) * | 2023-10-13 | 2024-01-05 | 安徽翼方健数信息科技有限公司 | Tumor targeting drug delivery treatment positioning needle system and assembly thereof |
| CN117352046B (en) * | 2023-10-13 | 2024-04-26 | 医数融和(安徽)科技有限公司 | Tumor targeting drug delivery treatment positioning needle system and assembly thereof |
| CN117457188A (en) * | 2023-11-08 | 2024-01-26 | 复旦大学附属中山医院 | System for evaluating histopathological changes in hepatocellular carcinoma caused by non-operative treatment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ke et al. | A neuro-heuristic approach for recognition of lung diseases from X-ray images | |
| Chen et al. | Artificial intelligence for assisting cancer diagnosis and treatment in the era of precision medicine | |
| Jiang et al. | An automatic detection system of lung nodule based on multigroup patch-based deep learning network | |
| Ker et al. | Automated brain histology classification using machine learning | |
| CN115527683A (en) | Method for predicting targeted therapeutic effect of Lunvatinib of liver cancer patient based on artificial intelligence | |
| Perez et al. | Automated lung cancer diagnosis using three-dimensional convolutional neural networks | |
| Li et al. | Molecular Subtypes Recognition of Breast Cancer in Dynamic Contrast‐Enhanced Breast Magnetic Resonance Imaging Phenotypes from Radiomics Data | |
| Jony et al. | Detection of lung cancer from CT scan images using GLCM and SVM | |
| Song et al. | Automatic coronary artery segmentation of CCTA images with an efficient feature-fusion-and-rectification 3D-UNet | |
| Avinash et al. | Early detection of malignant tumor in lungs using feed-forward neural network and k-nearest neighbor classifier | |
| Mitra et al. | Integrating radio imaging with gene expressions toward a personalized management of cancer | |
| CN113628761B (en) | Method for predicting treatment efficacy of anti-angiogenesis drug combined with immune checkpoint inhibitor | |
| Lin et al. | Using deep learning in ultrasound imaging of bicipital peritendinous effusion to grade inflammation severity | |
| CN116230215B (en) | Ovarian clear cell carcinoma prediction equipment, system and storable medium | |
| Yan et al. | Applicability analysis of immunotherapy for lung cancer patients based on deep learning | |
| Duan et al. | Segmentation of pulmonary vascular tree by incorporating vessel enhancement filter and variational region-growing | |
| NARİN et al. | The effect of hyper parameters on the classification of lung cancer images using deep learning methods | |
| Qin et al. | Classification for rectal CEUS images based on combining features by transfer learning | |
| Affane et al. | Robust deep 3-d architectures based on vascular patterns for liver vessel segmentation | |
| Sankar et al. | RETRACTED ARTICLE: Regression Neural Network segmentation approach with LIDC-IDRI for lung lesion | |
| Abdullah | GEOMETRICAL FEATURE OF LUNG LESION IDENTIFICATION USING COMPUTED TOMOGRAPHY SCAN IMAGES | |
| Wang et al. | Classification of metastatic hepatic carcinoma and hepatocellular carcinoma lesions using contrast‐enhanced CT based on EI‐CNNet | |
| Xiong et al. | Deep Ensemble Learning Network for Kidney Lesion Detection | |
| Braman | Novel Radiomics and Deep Learning Approaches Targeting the Tumor Environment to Predict Response to Chemotherapy | |
| Kulkarni et al. | Classification of liver cancer diseases using image processing technique |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20221227 |