Clinical analysis method and system for assisting chromosome AI karyotype analysis

Abstract

The invention discloses a clinical analysis method and system for assisting chromosome AI karyotype analysis. Relates to the technical field of chromosome detection. The method comprises the following steps: acquiring basic information, clinical information and chromosome karyotype analysis data of a subject, and preprocessing; extracting feature variables from the preprocessed data by using a feature selection method; constructing a clinical pre-analysis model based on the characteristic variables, performing preliminary analysis on clinical information of the subject to generate a pre-analysis result, and constructing a clinical scoring model according to the pre-analysis result; a karyotype analysis model is constructed, a preliminary karyotype report is output according to the pre-analysis result and the karyotype data as input of the karyotype analysis model, and the preliminary karyotype report comprises a karyotype classification result and a corresponding karyotype analysis chart; manually checking the chromosome karyotype preliminary report to obtain a final chromosome karyotype auditing report; and optimizing the clinical pre-analysis model and the karyotype analysis model according to the final chromosome karyotype auditing report.

Description

Technical Field

[0001] The present invention relates to the technical field of chromosome detection, and more specifically to a clinical analysis method and system for assisting chromosome AI karyotype analysis. Background Art

[0002] The chromosome karyotype analysis technology takes chromosomes in the metaphase of cell division as the research object. It compares, analyzes, sorts and numbers the chromosomes after banding according to the characteristics of chromosome length, band pattern, arm ratio, satellite, etc., and makes diagnosis based on the variation of chromosome structure and number. It has always been considered the "gold standard" for chromosome disease examination.

[0003] The chromosome karyotype analysis report generally includes basic case information, such as name, age, gender, consultation number, brief clinical history, metaphase map, karyotype map, collection date, report date, karyotype analysis opinion, test items (conventional G-banding in peripheral blood, high-resolution G-banding in peripheral blood, amniocentesis for prenatal diagnosis), etc.

[0004] Chromosome test reports are generally uploaded to the information system after three or four levels of proofreading. Although multiple manual proofreadings ensure a high accuracy rate, errors may still occur.

[0005] At present, the issuance of chromosome karyotype analysis reports cannot provide early warning for the above-mentioned information matching errors.

[0006] Therefore, proposing a clinical analysis method and system for assisting chromosome AI karyotype analysis is an issue that needs to be urgently addressed by those skilled in the art. Summary of the invention

[0007] In view of this, the present invention provides a clinical analysis method and system for assisted chromosome AI karyotype analysis, which can improve the accuracy and efficiency of diagnosis, optimize clinical decision-making, enhance the reliability and accuracy of the system, support remote review and collaboration, promote clinical research and quality control, and improve the overall performance of the medical system and the overall health level of patients.

[0008] In order to achieve the above object, the present invention adopts the following technical solution: A clinical analysis method for assisting chromosome AI karyotype analysis, comprising: Step 1: Obtain the basic information, clinical information and chromosome karyotype analysis data of the subjects and perform preprocessing; Step 2: Use feature selection method to extract feature variables from preprocessed data; Step 3: construct a clinical pre-analysis model based on the characteristic variables, conduct a preliminary analysis of the subject's clinical information, generate pre-analysis results, and construct a clinical scoring model based on the pre-analysis results, wherein the clinical scoring model is used to quantify the evaluation results; Step 4: constructing a karyotype analysis model, using the preliminary analysis results in combination with the chromosome karyotype data as input to the karyotype analysis model, and outputting a preliminary chromosome karyotype report, wherein the preliminary chromosome karyotype report includes a karyotype classification result and a corresponding chromosome karyotype analysis diagram; Step 5: Manually check the preliminary chromosome karyotype report to obtain the final chromosome karyotype review report; Step 6: Optimize the clinical pre-analysis model and karyotype analysis model according to the final chromosome karyotype review report.

[0009] Preferably, the basic information of the subject in step 1 includes the subject's name, gender, age, consultation number and consultation time; the clinical information includes the subject's specific symptoms, past medical history, and family genetic history; and the preprocessing method includes data cleaning, format conversion, and missing value filling.

[0010] Preferably, the clinical pre-analysis model includes a first input layer, a first feature extraction layer, a first classification decision layer and a first output layer; The first input layer is used to receive the clinical information of the subject; the first feature extraction layer uses a machine learning algorithm to extract features; the first classification decision layer intelligently classifies the subjects based on the extracted features; the first output layer gives the classification results and corresponding prediction probabilities of the subject's basic information, clinical information and chromosome karyotype analysis data.

[0011] Preferably, the clinical scoring model includes a second input layer, a second feature extraction layer, a scoring layer and a second output layer; the second input layer is used to receive the clinical information of the subject and the pre-analysis results; the second feature extraction layer uses a machine learning algorithm to extract clinical features; the scoring layer scores the subjects based on the extracted clinical features to assess their potential health risks or disease possibilities; the second output layer gives the scoring results and corresponding explanations.

[0012] Preferably, the karyotype analysis model comprises: a third input layer, a preprocessing layer, a feature learning layer, a second classification decision layer and a third output layer; The third input layer is used to receive the chromosome karyotype analysis data of the subject; the preprocessing layer cleans and converts the data; the feature learning layer uses a deep learning algorithm to automatically learn the feature representation of the chromosome karyotype; the second classification decision layer classifies the chromosome karyotype based on the feature representation and predicts the abnormal type of the chromosome karyotype; the third output layer outputs a preliminary report on the chromosome karyotype.

[0013] Preferably, based on the clinical preliminary analysis results and chromosome information of the subject, a preliminary chromosome karyotype report for manual verification is formed, and the preliminary chromosome karyotype report includes the subject's basic information, clinical information, chromosome information and preliminary karyotype analysis results; the subject's chromosome karyotype information in the preliminary karyotype analysis results is manually checked for consistency with the clinical information, and the karyotype analysis results are reviewed and corrected to obtain the subject's final chromosome karyotype review report.

[0014] Preferably, the process of optimizing the clinical pre-analysis model and the karyotype analysis model according to the final chromosome karyotype review report comprises: Collect and organize a large amount of chromosome karyotype analysis data and clinical information, and reconstruct the training data set based on the final chromosome karyotype review report; The training data set is trained and optimized through machine learning algorithms to re-obtain the clinical scoring model and karyotype analysis model; Integrate the trained model into the clinical analysis system for testing and verification; compare and analyze with actual clinical data.

[0015] The present invention also provides a clinical analysis system for assisting chromosome AI karyotype analysis, comprising: Preprocessing module: used to obtain the basic information, clinical information and chromosome karyotype analysis data of the subjects and perform preprocessing; Extraction module: extract feature variables from preprocessed data using feature selection methods; Model building module: used to build a clinical pre-analysis model based on the characteristic variables, conduct a preliminary analysis of the clinical information of the subject, generate pre-analysis results, build a clinical scoring model based on the pre-analysis results, and the clinical scoring model is used to quantify the evaluation results; build a karyotype analysis model, and output a preliminary chromosome karyotype report based on the pre-analysis results combined with chromosome karyotype data as input to the karyotype analysis model, and the preliminary chromosome karyotype report includes a karyotype classification result and a corresponding chromosome karyotype analysis diagram; Review module: used to manually review the preliminary chromosome karyotype report to obtain a final chromosome karyotype review report; Optimization module: optimizing the clinical pre-analysis model and the karyotype analysis model according to the final chromosome karyotype review report; The system is deployed on a cloud computing platform to provide cloud services for remote manual review. It can be seen from the above technical solution that, compared with the prior art, the present invention discloses a clinical analysis method and system for assisting chromosome AI karyotype analysis, which can effectively improve the accuracy and efficiency of diagnosis, optimize clinical decision-making, improve the reliability and accuracy of the system, support remote review and collaboration, promote clinical research and quality control; and improve the overall performance of the medical system and the overall health level of patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0017] Figure 1 A flow chart provided for the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] The present invention discloses a clinical analysis method for assisting chromosome AI karyotype analysis, see Figure 1 ,include: Step 1: Obtain the basic information, clinical information and chromosome karyotype analysis data of the subjects and perform preprocessing; Step 2: Use feature selection method to extract feature variables from preprocessed data; Step 3: construct a clinical pre-analysis model based on the characteristic variables, conduct a preliminary analysis of the subject's clinical information, generate pre-analysis results, and construct a clinical scoring model based on the pre-analysis results, wherein the clinical scoring model is used to quantify the evaluation results; Step 4: constructing a karyotype analysis model, using the preliminary analysis results in combination with the chromosome karyotype data as input to the karyotype analysis model, and outputting a preliminary chromosome karyotype report, wherein the preliminary chromosome karyotype report includes a karyotype classification result and a corresponding chromosome karyotype analysis diagram; Step 5: Manually check the preliminary chromosome karyotype report to obtain the final chromosome karyotype review report; Step 6: Optimize the clinical pre-analysis model and karyotype analysis model according to the final chromosome karyotype review report.

[0020] Specifically, the basic information of the subject in step 1 includes the subject's name, gender, age, consultation number and consultation time; the clinical information includes the subject's specific symptoms, past medical history, and family genetic history; the preprocessing method includes data cleaning, format conversion, and missing value filling.

[0021] Furthermore, in the embodiment of the present invention, after obtaining and processing the clinical information of the subject, the corresponding cases are collected according to the classification of the clinical information of the subject; after obtaining and processing the detailed clinical information of the subject, the subject is classified according to the key elements in this information (such as gender, age, purpose of medical treatment, disease type, symptom manifestation, past medical history and family genetic history, etc.), and then the corresponding cases are collected. The key to this process is to intelligently classify the clinical information of the subject, which is mainly based on the conventional standards for disease diagnosis, classification methods in medical literature and the experience judgment of clinical experts.

[0022] The types of genetic examinations obtained from the subjects can be divided into routine genetic examinations for pregnancy preparation, genetic examinations for assisted reproduction, genetic examinations related to child development, and genetic examinations related to tumors. The corresponding case data are collected according to the type of genetic examination.

[0023] The clinical information of the subjects is intelligently classified and a comprehensive clinical pre-analysis model is formed based on machine learning algorithms.

[0024] Specifically, the clinical pre-analysis model includes a first input layer, a first feature extraction layer, a first classification decision layer and a first output layer; The first input layer is used to receive the clinical information of the subject; the first feature extraction layer uses a machine learning algorithm to extract features; the first classification decision layer intelligently classifies the subjects based on the extracted features; the first output layer gives the classification results and corresponding prediction probabilities of the subject's basic information, clinical information and chromosome karyotype analysis data.

[0025] Furthermore, in the embodiments provided by the present invention, the collected clinical information is preprocessed, such as data cleaning, format conversion and missing value filling; feature selection technology is used to extract feature variables that have a significant impact on the classification results; based on these feature variables, a suitable machine learning model (such as a decision tree, support vector machine or neural network, etc.) is selected for training; the performance of the model is evaluated by cross-validation and other methods, and necessary parameter tuning is performed.

[0026] According to the obtained medical records and intelligent case classification, analysis was performed based on machine learning algorithms such as association rule analysis, weight factor analysis, and deep learning algorithms to construct the potential relationship between the clinical characteristics of the subjects and chromosome karyotype analysis.

[0027] The structure of the clinical pre-analysis model constructed in the embodiment enables the clinical pre-analysis model to efficiently and accurately process the clinical information of the subject, providing strong decision support for the next step of chromosome AI karyotype analysis and the final manual result review; and the clinical pre-analysis model can predict the number, structure and polymorphism of abnormal chromosomes that may exist in the subject and suitable chromosome detection items.

[0028] Specifically, the clinical scoring model includes a second input layer, a second feature extraction layer, a scoring layer and a second output layer; the second input layer is used to receive the clinical information of the subject and the pre-analysis results; the second feature extraction layer uses a machine learning algorithm to extract clinical features; the scoring layer scores the subject based on the extracted clinical features to assess their potential health risks or disease possibilities; the second output layer gives the scoring results and corresponding explanations.

[0029] Specifically, the karyotype analysis model includes: a third input layer, a preprocessing layer, a feature learning layer, a second classification decision layer and a third output layer; The third input layer is used to receive the chromosome karyotype analysis data of the subject; the preprocessing layer cleans and converts the data; the feature learning layer uses a deep learning algorithm to automatically learn the feature representation of the chromosome karyotype; the second classification decision layer classifies the chromosome karyotype based on the feature representation and predicts the abnormal type of the chromosome karyotype; the third output layer outputs a preliminary report on the chromosome karyotype.

[0030] Specifically, based on the clinical preliminary analysis results and chromosome information of the subject, a preliminary chromosome karyotype report for manual verification is formed, and the preliminary chromosome karyotype report includes the subject's basic information, clinical information, chromosome information and preliminary karyotype analysis results; the subject's chromosome karyotype information in the preliminary karyotype analysis results is manually checked for consistency with the clinical information, the karyotype analysis results are reviewed and corrected, and a final chromosome karyotype review report for the subject is obtained.

[0031] Specifically, the process of optimizing the clinical pre-analysis model and the karyotype analysis model according to the final chromosome karyotype review report includes: Collect and organize a large amount of chromosome karyotype analysis data and clinical information, and reconstruct the training data set based on the final chromosome karyotype review report; The training data set is trained and optimized through machine learning algorithms to re-obtain the clinical scoring model and karyotype analysis model; Integrate the trained model into the clinical analysis system for testing and verification; compare and analyze with actual clinical data.

[0032] Specifically, an embodiment of the present invention further provides a clinical analysis system for assisting chromosome AI karyotype analysis, comprising: Preprocessing module: used to obtain the basic information, clinical information and chromosome karyotype analysis data of the subjects and perform preprocessing; Extraction module: extract feature variables from preprocessed data using feature selection methods; Model building module: used to build a clinical pre-analysis model based on the characteristic variables, conduct a preliminary analysis of the clinical information of the subject, generate pre-analysis results, build a clinical scoring model based on the pre-analysis results, and the clinical scoring model is used to quantify the evaluation results; build a karyotype analysis model, and output a preliminary chromosome karyotype report based on the pre-analysis results combined with chromosome karyotype data as input to the karyotype analysis model, and the preliminary chromosome karyotype report includes a karyotype classification result and a corresponding chromosome karyotype analysis diagram; Review module: used to manually review the preliminary chromosome karyotype report to obtain a final chromosome karyotype review report; Optimization module: optimizing the clinical pre-analysis model and the karyotype analysis model according to the final chromosome karyotype review report; The system is deployed on a cloud computing platform to provide cloud services for remote manual review. In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0033] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A clinical analysis method for assisting chromosome AI karyotype analysis, characterized in that: include: Step 1: Obtain the basic information, clinical information and chromosome karyotype analysis data of the subjects and perform preprocessing; Step 2: Use feature selection method to extract feature variables from preprocessed data; Step 3: construct a clinical pre-analysis model based on the characteristic variables, conduct a preliminary analysis of the subject's clinical information, generate pre-analysis results, and construct a clinical scoring model based on the pre-analysis results. The clinical scoring model is used to quantify the evaluation results; Step 4: constructing a karyotype analysis model, and outputting a preliminary chromosome karyotype report based on the preliminary analysis results combined with the chromosome karyotype data as input to the karyotype analysis model, wherein the preliminary chromosome karyotype report includes a karyotype classification result and a corresponding chromosome karyotype analysis diagram; Step 5: Manually check the preliminary chromosome karyotype report to obtain the final chromosome karyotype review report; Step 6: Optimize the clinical pre-analysis model and karyotype analysis model according to the final chromosome karyotype review report.

2. A clinical analysis method for assisting chromosome AI karyotype analysis according to claim 1, characterized in that: The basic information of the subject in step 1 includes the subject's name, gender, age, consultation number and consultation time; the clinical information includes the subject's specific symptoms, past medical history, and family genetic history; the preprocessing methods include data cleaning, format conversion and missing value filling.

3. A clinical analysis method for assisting chromosome AI karyotype analysis according to claim 1, characterized in that: The clinical pre-analysis model includes a first input layer, a first feature extraction layer, a first classification decision layer and a first output layer; The first input layer is used to receive clinical information of the subject; the first feature extraction layer extracts features using a machine learning algorithm; The first classification decision layer intelligently classifies the subjects based on the extracted features; the first output layer gives the classification results and corresponding prediction probabilities of the subjects' basic information, clinical information and chromosome karyotype analysis data.

4. A clinical analysis method for assisting chromosome AI karyotype analysis according to claim 1, characterized in that: The clinical scoring model includes a second input layer, a second feature extraction layer, a scoring layer and a second output layer; the second input layer is used to receive the clinical information of the subject and the pre-analysis results; the second feature extraction layer uses a machine learning algorithm to extract clinical features; the scoring layer scores the subject based on the extracted clinical features to assess their potential health risks or disease possibilities; the second output layer gives the scoring results and corresponding explanations.

5. A clinical analysis method for assisting chromosome AI karyotype analysis according to claim 1, characterized in that: The karyotype analysis model includes: a third input layer, a preprocessing layer, a feature learning layer, a second classification decision layer and a third output layer; The third input layer is used to receive the chromosome karyotype analysis data of the subject; the preprocessing layer cleans and converts the data; the feature learning layer uses a deep learning algorithm to automatically learn the feature representation of the chromosome karyotype; the second classification decision layer classifies the chromosome karyotype based on the feature representation and predicts the abnormal type of the chromosome karyotype; the third output layer outputs a preliminary report on the chromosome karyotype.

6. A clinical analysis method for assisting chromosome AI karyotype analysis according to claim 1, characterized in that: According to the clinical preliminary analysis results and chromosome information of the subject, a preliminary chromosome karyotype report for manual verification is formed, wherein the preliminary chromosome karyotype report includes the basic information, clinical information, chromosome information and preliminary karyotype analysis results of the subject; The consistency between the subject's chromosome karyotype information and clinical information in the preliminary karyotype analysis results is manually checked, and the karyotype analysis results are reviewed and corrected to obtain the subject's final chromosome karyotype review report.

7. A clinical analysis method for assisting chromosome AI karyotype analysis according to claim 1, characterized in that: The process of optimizing the clinical pre-analysis model and the karyotype analysis model based on the final karyotype review report includes: Collect and organize a large amount of chromosome karyotype analysis data and clinical information, and reconstruct the training data set based on the final chromosome karyotype review report; The training data set is trained and optimized through machine learning algorithms to re-obtain the clinical scoring model and karyotype analysis model; Integrate the trained model into the clinical analysis system for testing and verification; compare and analyze with actual clinical data.

8. A clinical analysis system for assisting chromosome AI karyotype analysis, characterized in that: include: Preprocessing module: used to obtain the basic information, clinical information and chromosome karyotype analysis data of the subjects and perform preprocessing; Extraction module: extract feature variables from preprocessed data using feature selection methods; Model building module: used to build a clinical pre-analysis model based on the characteristic variables, conduct a preliminary analysis of the clinical information of the subject, generate pre-analysis results, build a clinical scoring model based on the pre-analysis results, and the clinical scoring model is used to quantify the evaluation results; build a karyotype analysis model, and output a preliminary chromosome karyotype report based on the pre-analysis results combined with chromosome karyotype data as input to the karyotype analysis model, and the preliminary chromosome karyotype report includes a karyotype classification result and a corresponding chromosome karyotype analysis diagram; Review module: used to manually review the preliminary chromosome karyotype report to obtain a final chromosome karyotype review report; Optimization module: optimizing the clinical pre-analysis model and the karyotype analysis model according to the final chromosome karyotype review report; The system is deployed on a cloud computing platform to provide cloud services for remote manual review.

Images