A clinical research management method for chronic respiratory diseases

Abstract

The present invention discloses a clinical research management method for chronic respiratory diseases, and relates to the field of medical management technology. A clinical research management method for chronic respiratory diseases, comprising the following steps: clinical research information data collection for chronic respiratory diseases, clinical research typing for chronic respiratory diseases, clinical research scheme matching for chronic respiratory diseases, and clinical research management for chronic respiratory diseases. The present invention uses a clinical main feature analysis model for chronic respiratory diseases, which is used to use a main feature extraction algorithm that combines an improved principal component analysis method and a clustering algorithm to analyze the main feature data and initial labels for each chronic respiratory disease patient data, and can effectively extract key information from a large amount of patient data, improve the speed and accuracy of data processing, and more accurately identify the patient's condition, thereby improving the diagnostic quality of clinical research; by using a clinical research classification model for chronic respiratory diseases, the characteristics of each patient's condition can be more accurately identified, which helps to improve the pertinence and effectiveness of treatment.

Description

Technical Field

[0001] The present invention relates to the field of medical management technology, and in particular to a clinical research management method for chronic respiratory diseases. Background Art

[0002] Chronic respiratory diseases refer to chronic respiratory diseases. Common chronic respiratory diseases include chronic obstructive pulmonary disease, asthma, pulmonary hypertension, etc. In the existing clinical research management methods for chronic respiratory diseases, chronic respiratory diseases such as asthma usually require long-term management. Existing research may not fully consider the chronicity and complexity of the disease and cannot provide continuous medical follow-up services to patients with chronic respiratory diseases. At the same time, because the symptoms and reactions of different patients with chronic respiratory diseases may vary greatly, a unified treatment plan may not be suitable for all patients and does not fully consider individual differences in patients, such as age, gender, and lifestyle habits. Patients with chronic respiratory diseases may not have enough opportunities to participate in the decision-making process of clinical research, resulting in patients' needs and preferences not being fully reflected. Summary of the invention

[0003] The present invention aims to provide a clinical research management method for chronic respiratory diseases, so as to provide the most accurate clinical research classification and personalized treatment plans for different chronic respiratory disease patients.

[0004] A method for clinical research management of chronic respiratory diseases, comprising the following steps:

[0005] S1. Data collection of clinical research information on chronic respiratory diseases

[0006] Get the current respiratory chronic disease patient information data set to be grouped; the current respiratory chronic disease patient information data set to be grouped contains N respiratory chronic disease patient data D n , n=1, 2, ..., N, each piece of respiratory chronic disease patient data D n Including patients with chronic respiratory diseases n Corresponding respiratory chronic disease patient information and respiratory chronic disease patient medical data information; input the current respiratory chronic disease patient information data set to be grouped into the respiratory chronic disease clinical main feature analysis model for initial analysis, and obtain the main feature data Z of the respiratory chronic disease patient n and the initial classification label F n The clinical main feature analysis model of chronic respiratory diseases includes a data preprocessing layer, a data optimization clustering layer, a main component analysis layer for chronic respiratory diseases and a main feature output layer, which is used to analyze the main feature data and initial classification labels for each chronic respiratory disease patient data using a main feature extraction algorithm that combines an improved principal component analysis method with a clustering algorithm;

[0007] S2. Clinical research classification of chronic respiratory diseases

[0008] The main feature data Z of patients with chronic respiratory diseases nand the initial classification label F n Input into the classification model of clinical research on chronic respiratory diseases for analysis, and obtain the classification result of chronic respiratory diseases patients B n ; According to the classification results of patients with chronic respiratory diseases B n Match the corresponding initial chronic respiratory disease patient tracking treatment plan A n The classification model for clinical research on chronic respiratory diseases includes a data feature fusion layer, a data feature classification layer, and a classification result output layer, which is used to match the classification results of chronic respiratory disease patients using feature fusion and feature recognition algorithms. n ;

[0009] S3. Matching of clinical research plans for chronic respiratory diseases

[0010] According to the initial chronic respiratory disease patient follow-up treatment plan A n Assign respiratory chronic disease clinical research manager Y n ; Collect P of patients with chronic respiratory diseases n and Y, clinical research manager for chronic respiratory diseases n Decision data, get the dynamic decision data set V n ; Dynamic decision dataset V n Follow-up treatment plan A for patients with initial respiratory chronic diseases n Input into the dynamic decision-making model for clinical research on chronic respiratory diseases for analysis, and obtain the optimized follow-up treatment plan A for patients with chronic respiratory diseases n '; The dynamic decision-making model for clinical research on chronic respiratory diseases includes a text feature extraction layer, a differential dynamic factor analysis layer, and an optimization solution output layer, which is used to combine the P values ​​of patients with chronic respiratory diseases n and Y, clinical research manager for chronic respiratory diseases n Decision-making data for initial chronic respiratory disease patients to track treatment plans A n Make optimization adjustments;

[0011] S4. Clinical research management of chronic respiratory diseases

[0012] Repeat steps S2 and S3 to traverse the main feature data Z of all patients with chronic respiratory diseases n Until all the chronic respiratory disease patient data D in the current chronic respiratory disease patient information data set to be grouped n Match the corresponding optimized respiratory chronic disease patient tracking treatment plan A n ', will optimize the follow-up treatment plan for patients with chronic respiratory diseases n 'Store in the respiratory chronic disease clinical research management database to complete a respiratory chronic disease clinical research management.

[0013] As a preferred technical solution of the present invention, the clinical main feature analysis model of chronic respiratory diseases in step S1 includes a data preprocessing layer, a data optimization clustering layer, a main component analysis layer of chronic respiratory diseases and a main feature output layer;

[0014] The data preprocessing layer is used to process all the chronic respiratory disease patient data D in the current chronic respiratory disease patient information data set to be grouped. n Perform feature extraction to obtain the data features T of patients with chronic respiratory diseases n , T n ={T n1 , T n2 ,…,T nm ,…,T nM}, m=1, 2, ..., M, M represents the data characteristics of patients with chronic respiratory diseases T n The total number of different characteristic components in ;

[0015] The data optimization clustering layer is used to classify the data features T of all patients with chronic respiratory diseases n The optimized kernel fuzzy C-means clustering algorithm with Gaussian radial basis function as kernel function is used for clustering to obtain the initial classification label F n ;

[0016] The principal component analysis layer of respiratory chronic diseases is used to analyze the data features of all patients with respiratory chronic diseases. n Perform principal component analysis to obtain the main characteristic data Z of patients with chronic respiratory diseases n ;

[0017] The main feature output layer outputs the main feature data Z of patients with chronic respiratory diseases n And the corresponding initial classification label F n .

[0018] As a preferred technical solution of the present invention, the specific steps of clustering in the data optimization clustering layer include:

[0019] Construct K dynamic cluster center individuals G k , k=1, 2, ..., K; each dynamic cluster center individual G k Contains the solution of clustering centers for optimizing the kernel fuzzy C-means clustering algorithm; K dynamic clustering center individuals G k Combine to obtain the dynamic clustering iterative population of chronic respiratory diseases; set the maximum number of iterations;

[0020] Dynamic clustering center individual G in iterative population based on dynamic clustering of chronic respiratory diseases k Data characteristics of patients with chronic respiratory diseases T n Clustering is performed using the optimized kernel fuzzy C-means clustering algorithm with Gaussian radial basis function as the kernel function to obtain simulated clustering results; clustering effect evaluation is performed on the simulated clustering results to obtain the clustering effect evaluation result L k ; The clustering effect evaluation result L k As the dynamic cluster center individual G kThe fitness S k ;

[0021] When the maximum number of iterations is reached, the dynamic clustering center individual G with the largest fitness in the dynamic clustering iteration population of chronic respiratory diseases is output. k , which is the optimal dynamic clustering center individual; the solution of the clustering center in the optimal dynamic clustering center individual is used as the solution of the initial classification label, and the data feature T of all respiratory chronic disease patients after clustering is n Match the initial classification label F n .

[0022] As a preferred technical solution of the present invention, the classification model for clinical research on chronic respiratory diseases in step S2 includes a data feature fusion layer, a data feature classification layer and a classification result output layer;

[0023] The data feature fusion layer is used to transform the main feature data Z of patients with chronic respiratory diseases n and the initial classification label F n Perform feature fusion to obtain the fusion feature R of patients with chronic respiratory diseases n ;

[0024] The data feature classification layer is used to fuse the features R according to the patients with chronic respiratory diseases. n Perform feature analysis to obtain the classification results of patients with chronic respiratory diseases B n ;

[0025] The classification result output layer is used to output the classification results of patients with chronic respiratory diseases B n .

[0026] As a preferred technical solution of the present invention, the specific steps of training the data feature classification layer include:

[0027] Collect several groups of respiratory chronic disease typing matching training samples; each group of respiratory chronic disease typing matching training samples contains a group of historical respiratory chronic disease patient fusion features and corresponding respiratory chronic disease patient typing results, and the corresponding respiratory chronic disease patient typing results are used as the target value of each group of respiratory chronic disease typing matching training samples; combine several groups of respiratory chronic disease typing matching training samples to obtain a respiratory chronic disease typing matching training set;

[0028] The respiratory chronic disease typing matching training set is input into the respiratory chronic disease clinical research classification model to train the data feature classification layer with the target value as the target to obtain the initial data feature classification layer; the initial data feature classification layer is model evaluated to obtain the initial data feature classification layer model evaluation result; if the initial data feature classification layer model evaluation result is passed, the trained initial data feature classification layer is used as the data feature classification layer in the respiratory chronic disease clinical research classification model; otherwise, the respiratory chronic disease typing matching training set is used to continue model training.

[0029] As a preferred technical solution of the present invention, according to the classification results of chronic respiratory disease patients B n Match the corresponding initial chronic respiratory disease patient tracking treatment plan A n The specific steps include:

[0030] Obtain a standard library of tracking and treatment plans for patients with chronic respiratory diseases, which contains several groups of standard chronic respiratory disease patient classification results and corresponding standard chronic respiratory disease patient tracking and treatment plans;

[0031] The results of the classification of patients with chronic respiratory diseases are B n Match the most similar standard respiratory chronic disease patient classification result in the standard database of respiratory chronic disease patient tracking treatment plans, and use the standard respiratory chronic disease patient tracking treatment plan corresponding to the most similar standard respiratory chronic disease patient classification result as respiratory chronic disease patient classification result B. n Corresponding initial chronic respiratory disease patient follow-up treatment plan A n .

[0032] As a preferred technical solution of the present invention, the dynamic decision-making model for clinical research on chronic respiratory diseases in step S3 includes a text feature extraction layer, a difference dynamic factor analysis layer and an optimization solution output layer;

[0033] The text feature extraction layer is used to extract the dynamic decision dataset V n Extract features from the decision data in the dynamic decision data feature set V n ',V n '={V n1 , V n2 , …, V nj , …, V nJ}, J is the dynamic decision data feature set V n 'The number of decision data item features, V nj Represents the dynamic decision data feature set V n 'The decision data item characteristics in ';

[0034] The differential dynamic factor analysis layer is used to analyze the dynamic decision data feature set V n 'Perform feature recognition and obtain the decision data difference factor set C n , C n ={C n1 , C n2 , …, C nj , …, C nJ}, C nj Represents the sub-item features of decision data V nj The corresponding matching decision data difference factor;

[0035] The optimization solution output layer is used to calculate the difference factor set C based on the decision data. n Follow-up treatment plan A for patients with initial chronic respiratory diseases n Adjustments were made to obtain an optimized follow-up treatment plan for patients with chronic respiratory diseases A n '.

[0036] As a preferred technical solution of the present invention, the specific steps of training the difference dynamic factor analysis layer include:

[0037] Collecting several groups of dynamic difference factor training samples, each group of dynamic difference factor training samples includes a group of prior dynamic decision data features and corresponding difference dynamic factors; using the corresponding difference dynamic factors as target values ​​for each group of dynamic difference factor training samples; combining several groups of dynamic difference factor training samples to obtain a dynamic difference factor training set;

[0038] The dynamic differential factor training set is input into the dynamic decision-making model for clinical research on chronic respiratory diseases, and the differential dynamic factor analysis layer is trained with the target value as the target to obtain the initial differential dynamic factor analysis layer; the initial differential dynamic factor analysis layer is evaluated to obtain the model evaluation result of the initial differential dynamic factor analysis layer; if the model evaluation result of the initial differential dynamic factor analysis layer is passed, the trained initial differential dynamic factor analysis layer is used as the differential dynamic factor analysis layer in the dynamic decision-making model for clinical research on chronic respiratory diseases; otherwise, the dynamic differential factor training set is used to continue model training.

[0039] The present invention has the following advantages:

[0040] 1. The present invention adopts a clinical main feature analysis model for chronic respiratory diseases, and uses a main feature extraction algorithm that combines an improved principal component analysis method with a clustering algorithm to analyze the main feature data and initial labels for each chronic respiratory disease patient data. This can effectively extract key information from a large amount of patient data, reduce unnecessary data redundancy, and improve the speed and accuracy of data processing. At the same time, it can more accurately identify the patient's condition, thereby improving the diagnostic quality of clinical research. The clinical research classification model for chronic respiratory diseases uses feature fusion and feature recognition algorithms to analyze the main feature data and initial labels for each chronic respiratory disease patient data to match the chronic respiratory disease patient typing results. This can more accurately identify the characteristics of each patient's condition, thereby realizing individualized typing management, which helps to improve the pertinence and effectiveness of treatment.

[0041] 2. The present invention uses a dynamic decision-making model for clinical research on chronic respiratory diseases combined with the P n and Y, clinical research manager for chronic respiratory diseases n Decision-making data for initial chronic respiratory disease patients to track treatment plans A nOptimization and adjustment will help to continuously optimize the patient's follow-up treatment plan, strengthen communication between doctors and patients, ensure that patients can better understand their condition and treatment plan, improve treatment compliance, and ensure that treatment strategies are adjusted in time as the patient's condition changes, thereby improving the success rate of treatment and patient satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 The present invention is a flowchart of a method for clinical research management of chronic respiratory diseases adopted in an embodiment of the present invention. DETAILED DESCRIPTION

[0043] In order to enable persons skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0044] A management method for clinical research on chronic respiratory diseases, see Figure 1 As shown, the following steps are included:

[0045] S1. Data collection of clinical research information on chronic respiratory diseases

[0046] Get the current respiratory chronic disease patient information data set to be grouped; the current respiratory chronic disease patient information data set to be grouped contains N respiratory chronic disease patient data D n , n=1, 2, ..., N, each piece of respiratory chronic disease patient data D n Including patients with chronic respiratory diseases n Corresponding respiratory chronic disease patient information and respiratory chronic disease patient medical data information; input the current respiratory chronic disease patient information data set to be grouped into the respiratory chronic disease clinical main feature analysis model for initial analysis, and obtain the main feature data Z of the respiratory chronic disease patient n and the initial classification label F n The clinical main feature analysis model of chronic respiratory diseases includes a data preprocessing layer, a data optimization clustering layer, a main component analysis layer for chronic respiratory diseases and a main feature output layer, which is used to analyze the main feature data and initial classification labels for each chronic respiratory disease patient data using a main feature extraction algorithm that combines an improved principal component analysis method with a clustering algorithm;

[0047] When conducting clinical research management of chronic respiratory diseases, it is necessary to dynamically update the clinical research management database of chronic respiratory diseases to improve the real-time update of disease data and synchronous update of treatment plans during the tracking of chronic respiratory diseases; the main feature extraction algorithm combined with the improved principal component analysis and clustering algorithm can be used to extract the key information that best represents the patient's characteristics from complex patient information and medical data, which makes the subsequent analysis more focused on the important factors affecting the disease process; by automatically generating initial classification labels, the workload of manual classification is reduced and the classification efficiency is improved. In addition, since it is a data-driven method, it can theoretically reduce the influence of subjective bias and make classification more objective; through accurate feature extraction and initial classification, more personalized medical services can be provided to each patient, which helps to improve treatment effects and reduce unnecessary treatment side effects;

[0048] The clinical main feature analysis model of chronic respiratory diseases in step S1 includes a data preprocessing layer, a data optimization clustering layer, a main component analysis layer of chronic respiratory diseases and a main feature output layer;

[0049] The data preprocessing layer is used to process all the chronic respiratory disease patient data D in the current chronic respiratory disease patient information data set to be grouped. n Perform feature extraction to obtain the data features T of patients with chronic respiratory diseases n , T n ={T n1 , T n2 ,…,T nm ,…,T nM}, m=1, 2, ..., M, M represents the data characteristics of patients with chronic respiratory diseases T n The total number of different characteristic components in ;

[0050] The data optimization clustering layer is used to classify the data features T of all patients with chronic respiratory diseases n The optimized kernel fuzzy C-means clustering algorithm with Gaussian radial basis function as kernel function is used for clustering to obtain the initial classification label F n ;

[0051] The principal component analysis layer of respiratory chronic diseases is used to analyze the data features of all patients with respiratory chronic diseases. n Perform principal component analysis to obtain the main characteristic data Z of patients with chronic respiratory diseases n ;

[0052] The main feature output layer outputs the main feature data Z of patients with chronic respiratory diseases n And the corresponding initial classification label F n ;

[0053] The specific steps of clustering in the data optimization clustering layer include:

[0054] Construct K dynamic cluster center individuals Gk , k=1, 2, ..., K; each dynamic cluster center individual G k Contains the solution of clustering centers for optimizing the kernel fuzzy C-means clustering algorithm; K dynamic clustering center individuals G k Combine to obtain the dynamic clustering iterative population of chronic respiratory diseases; set the maximum number of iterations;

[0055] Dynamic clustering center individual G in iterative population based on dynamic clustering of chronic respiratory diseases k Data characteristics of patients with chronic respiratory diseases T n Clustering is performed using the optimized kernel fuzzy C-means clustering algorithm with Gaussian radial basis function as the kernel function to obtain simulated clustering results; clustering effect evaluation is performed on the simulated clustering results to obtain the clustering effect evaluation result L k ; The clustering effect evaluation result L k As the dynamic cluster center individual G k The fitness S k ;

[0056] When the maximum number of iterations is reached, the dynamic clustering center individual G with the largest fitness in the dynamic clustering iteration population of chronic respiratory diseases is output. k , which is the optimal dynamic clustering center individual; the solution of the clustering center in the optimal dynamic clustering center individual is used as the solution of the initial classification label, and the data feature T of all respiratory chronic disease patients after clustering is n Match the initial classification label F n ;

[0057] The optimized kernel fuzzy C-means clustering algorithm using Gaussian radial basis function as kernel function can effectively cluster complex data. This method not only considers the distance between data points, but also the nonlinear relationship between them, thereby improving the clustering accuracy; constructing dynamic clustering center individual G k And form an iterative population, allowing the cluster center to be continuously adjusted during the iteration process until the optimal solution is found, which means that the final clustering result is closer to the actual data distribution and improves the accuracy of classification; the principal component analysis layer of chronic respiratory diseases can help reduce the data dimension while retaining the main variation information in the data, which helps to simplify the data set without losing important information, facilitating subsequent analysis and visualization; combining the particle swarm iteration algorithm with the clustering algorithm can expand the search range of the cluster center, avoid the clustering algorithm from falling into the local optimal solution, and find a better cluster center;

[0058] S2. Clinical research classification of chronic respiratory diseases

[0059] The main feature data Z of patients with chronic respiratory diseases n and the initial classification label F nInput into the classification model of clinical research on chronic respiratory diseases for analysis, and obtain the classification result of chronic respiratory diseases patients B n ; According to the classification results of patients with chronic respiratory diseases B n Match the corresponding initial chronic respiratory disease patient tracking treatment plan A n The classification model for clinical research on chronic respiratory diseases includes a data feature fusion layer, a data feature classification layer, and a classification result output layer, which is used to match the classification results of chronic respiratory disease patients using feature fusion and feature recognition algorithms. n ;

[0060] According to the classification results of patients with chronic respiratory diseases B n Match the corresponding initial chronic respiratory disease patient tracking treatment plan A n The specific steps include:

[0061] Obtain a standard library of tracking and treatment plans for patients with chronic respiratory diseases, which contains several groups of standard chronic respiratory disease patient classification results and corresponding standard chronic respiratory disease patient tracking and treatment plans;

[0062] The results of the classification of patients with chronic respiratory diseases are B n Match the most similar standard respiratory chronic disease patient classification result in the standard database of respiratory chronic disease patient tracking treatment plans, and use the standard respiratory chronic disease patient tracking treatment plan corresponding to the most similar standard respiratory chronic disease patient classification result as respiratory chronic disease patient classification result B. n Corresponding initial chronic respiratory disease patient follow-up treatment plan A n The standard database for tracking treatment plans for patients with chronic respiratory diseases is established through a large amount of historical data, which conforms to the scientificity and standardization of clinical research and provides a basis for personalized treatment plans for different patients.

[0063] The classification model for clinical research on chronic respiratory diseases in step S2 includes a data feature fusion layer, a data feature classification layer and a classification result output layer;

[0064] The data feature fusion layer is used to transform the main feature data Z of patients with chronic respiratory diseases n and the initial classification label F n Perform feature fusion to obtain the fusion feature R of patients with chronic respiratory diseases n ;

[0065] The data feature classification layer is used to fuse the features R according to the patients with chronic respiratory diseases. n Perform feature analysis to obtain the classification results of patients with chronic respiratory diseases B n ;

[0066] The classification result output layer is used to output the classification results of patients with chronic respiratory diseases B n ;

[0067] The specific steps of training data feature classification layer include:

[0068] Collect several groups of respiratory chronic disease typing matching training samples; each group of respiratory chronic disease typing matching training samples contains a group of historical respiratory chronic disease patient fusion features and corresponding respiratory chronic disease patient typing results, and the corresponding respiratory chronic disease patient typing results are used as the target value of each group of respiratory chronic disease typing matching training samples; combine several groups of respiratory chronic disease typing matching training samples to obtain a respiratory chronic disease typing matching training set;

[0069] Input the respiratory chronic disease classification matching training set into the respiratory chronic disease clinical research classification model to train the data feature classification layer with the target value as the target, and obtain the initial data feature classification layer; perform model evaluation on the initial data feature classification layer to obtain the initial data feature classification layer model evaluation result; if the initial data feature classification layer model evaluation result is passed, the trained initial data feature classification layer is used as the data feature classification layer in the respiratory chronic disease clinical research classification model; otherwise, continue model training using the respiratory chronic disease classification matching training set;

[0070] Through feature fusion and feature recognition algorithms, the main feature data and initial classification labels of patients can be analyzed more comprehensively, so as to obtain more accurate classification results. This precise classification helps to better understand the disease type and progression of patients; matching the corresponding initial treatment plan according to the classification results ensures that each patient can get the treatment plan that best suits their condition, which helps to improve the effectiveness of treatment and the speed of patient recovery; using historical data for feature fusion and classification training can provide data support for clinical decision-making, making the selection of treatment plans more scientific and reasonable, and reducing the uncertainty of subjective judgment; using the respiratory chronic disease classification matching training set to train the data feature classification layer, and ensuring the effectiveness of the model through model evaluation, which ensures the reliability and generalization ability of the model; determining the treatment plan by matching the standard classification results with the greatest similarity ensures the consistency and standardization of the treatment plan, which helps to improve the overall treatment level and service quality;

[0071] S3. Matching of clinical research plans for chronic respiratory diseases

[0072] According to the initial chronic respiratory disease patient follow-up treatment plan A n Assign respiratory chronic disease clinical research manager Y n ; Collect P of patients with chronic respiratory diseases n and Y, clinical research manager for chronic respiratory diseases n Decision data, get the dynamic decision data set V n ; Dynamic decision dataset V n Follow-up treatment plan A for patients with initial respiratory chronic diseases nInput into the dynamic decision-making model for clinical research on chronic respiratory diseases for analysis, and obtain the optimized follow-up treatment plan A for patients with chronic respiratory diseases n '; The dynamic decision-making model for clinical research on chronic respiratory diseases includes a text feature extraction layer, a differential dynamic factor analysis layer, and an optimization solution output layer, which is used to combine the P values ​​of patients with chronic respiratory diseases n and Y, clinical research manager for chronic respiratory diseases n Decision-making data for initial chronic respiratory disease patients to track treatment plans A n Make optimization adjustments;

[0073] The dynamic decision-making model for clinical research on chronic respiratory diseases in step S3 includes a text feature extraction layer, a difference dynamic factor analysis layer, and an optimization solution output layer;

[0074] The text feature extraction layer is used to extract the dynamic decision dataset V n Extract features from the decision data in the dynamic decision data feature set V n ',V n '={V n1 , V n2 , …, V nj , …, V nJ}, J is the dynamic decision data feature set V n 'The number of decision data item features, V nj Represents the dynamic decision data feature set V n 'The decision data item characteristics in ';

[0075] The differential dynamic factor analysis layer is used to analyze the dynamic decision data feature set V n 'Perform feature recognition and obtain the decision data difference factor set C n , C n ={C n1 , C n2 , …, C nj , …, C nJ}, C nj Represents the sub-item features of decision data V nj The corresponding matching decision data difference factor;

[0076] The optimization solution output layer is used to calculate the difference factor set C based on the decision data. n Follow-up treatment plan A for patients with initial chronic respiratory diseases n Adjustments were made to obtain an optimized follow-up treatment plan for patients with chronic respiratory diseases A n '; Based on the decision data difference factor set C n The decision data difference factor in the initial respiratory chronic disease patient follow-up treatment plan A n Make minor adjustments and changes to the various related plans in order to meet the ultimate needs of the patient;

[0077] The specific steps for training the differential dynamic factor analysis layer include:

[0078] Collecting several groups of dynamic difference factor training samples, each group of dynamic difference factor training samples includes a group of prior dynamic decision data features and corresponding difference dynamic factors; using the corresponding difference dynamic factors as target values ​​for each group of dynamic difference factor training samples; combining several groups of dynamic difference factor training samples to obtain a dynamic difference factor training set;

[0079] The dynamic differential factor training set is input into the dynamic decision-making model for clinical research on chronic respiratory diseases to train the differential dynamic factor analysis layer with the target value as the target, and obtain the initial differential dynamic factor analysis layer; the model evaluation of the initial differential dynamic factor analysis layer is performed to obtain the model evaluation result of the initial differential dynamic factor analysis layer; if the model evaluation result of the initial differential dynamic factor analysis layer is passed, the trained initial differential dynamic factor analysis layer is used as the differential dynamic factor analysis layer in the dynamic decision-making model for clinical research on chronic respiratory diseases; otherwise, the model training is continued using the dynamic differential factor training set;

[0080] By analyzing the decision-making data of chronic respiratory disease patients and clinical research managers, we can find the deficiencies in the treatment plan and optimize it accordingly, so that the final treatment plan is more in line with the actual needs of the patient; the dynamic decision-making model can dynamically adjust the treatment plan according to the latest decision-making data. This flexibility ensures that the treatment plan can respond to changes in the patient's condition in a timely manner and improve the treatment effect; by extracting features from the decision data through the text feature extraction layer, we can more systematically analyze the decision-making behavior of patients and managers and provide data support for subsequent optimization; the differential dynamic factor analysis layer can identify the key factors that affect the treatment effect, helping clinicians and researchers to more accurately grasp which aspects need to be adjusted, so as to achieve more effective treatment; collecting decision-making data from both doctors and patients will help enhance cooperation between the two parties, jointly participate in the formulation of treatment plans, and improve patient satisfaction;

[0081] S4. Clinical research management of chronic respiratory diseases

[0082] Repeat steps S2 and S3 to traverse the main feature data Z of all patients with chronic respiratory diseases n Until all the chronic respiratory disease patient data D in the current chronic respiratory disease patient information data set to be grouped n Match the corresponding optimized respiratory chronic disease patient tracking treatment plan A n ', will optimize the follow-up treatment plan for patients with chronic respiratory diseases n'Store it in the clinical research management library for chronic respiratory diseases to complete the clinical research management of chronic respiratory diseases; by repeating steps S2 and S3, ensure that the main characteristic data of all patients are analyzed, and formulate an optimized follow-up treatment plan for each patient. This management process improves the organization and efficiency of the entire clinical research; store the optimized treatment plan in the clinical research management library to ensure the standardization and traceability of the plan, which is convenient for subsequent research and reference.

[0083] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all such improvements and changes should fall within the scope of protection of the appended claims of the present invention. Parts not described in detail in this specification belong to the prior art known to those skilled in the art.

Claims

1. A method for clinical research management of chronic respiratory diseases, characterized in that: The following steps are involved: S1. Data collection of clinical research information on chronic respiratory diseases Get the current respiratory chronic disease patient information data set to be grouped; the current respiratory chronic disease patient information data set to be grouped contains N respiratory chronic disease patient data D n , n=1, 2, ..., N, each piece of respiratory chronic disease patient data D n Including patients with chronic respiratory diseases n Corresponding information of patients with chronic respiratory diseases and medical data information of patients with chronic respiratory diseases; The current respiratory chronic disease patient information data set to be grouped is input into the respiratory chronic disease clinical main feature analysis model for initial analysis to obtain the main feature data Z of the respiratory chronic disease patients. n and the initial classification label F n ; The clinical main feature analysis model of chronic respiratory diseases includes a data preprocessing layer, a data optimization clustering layer, a main component analysis layer for chronic respiratory diseases, and a main feature output layer. It is used to analyze the main feature data and initial classification labels for each chronic respiratory disease patient data using a main feature extraction algorithm that combines an improved principal component analysis method with a clustering algorithm. S2. Clinical research classification of chronic respiratory diseases The main feature data Z of patients with chronic respiratory diseases n and the initial classification label F n Input into the classification model of clinical research on chronic respiratory diseases for analysis, and obtain the classification result of chronic respiratory diseases patients B n ; According to the classification results of patients with chronic respiratory diseases B n Match the corresponding initial chronic respiratory disease patient tracking treatment plan A n ; The classification model for clinical research on chronic respiratory diseases includes a data feature fusion layer, a data feature classification layer, and a classification result output layer, which is used to match the classification results of chronic respiratory disease patients using feature fusion and feature recognition algorithms. n ; S3. Matching of clinical research plans for chronic respiratory diseases According to the initial chronic respiratory disease patient follow-up treatment plan A n Assign respiratory chronic disease clinical research manager Y n ; Collect P of patients with chronic respiratory diseases n and Y, clinical research manager for chronic respiratory diseases n Decision data, get the dynamic decision data set V n ; Dynamic decision dataset V n Follow-up treatment plan A for patients with initial respiratory chronic diseases n Input into the dynamic decision-making model for clinical research on chronic respiratory diseases for analysis, and obtain the optimized follow-up treatment plan A for patients with chronic respiratory diseases n '; The dynamic decision-making model for clinical research on chronic respiratory diseases includes a text feature extraction layer, a differential dynamic factor analysis layer, and an optimization solution output layer. n and Y, clinical research manager for chronic respiratory diseases n Decision-making data for initial chronic respiratory disease patients to track treatment plans A n Make optimization adjustments; S4. Clinical research management of chronic respiratory diseases Repeat steps S2 and S3 to traverse the main feature data Z of all patients with chronic respiratory diseases n Until all the chronic respiratory disease patient data D in the current chronic respiratory disease patient information data set to be grouped n Match the corresponding optimized respiratory chronic disease patient tracking treatment plan A n ', will optimize the follow-up treatment plan for patients with chronic respiratory diseases n 'Store in the respiratory chronic disease clinical research management database to complete a respiratory chronic disease clinical research management; The clinical main feature analysis model of chronic respiratory diseases in step S1 includes a data preprocessing layer, a data optimization clustering layer, a main component analysis layer of chronic respiratory diseases and a main feature output layer; The data preprocessing layer is used to process all the chronic respiratory disease patient data D in the current chronic respiratory disease patient information data set to be grouped. n Perform feature extraction to obtain the data features T of patients with chronic respiratory diseases n , T n ={T n1 , T n2 ,…,T nm ,…,T nM }, m=1, 2, ..., M, M represents the data characteristics of patients with chronic respiratory diseases T n The total number of different characteristic components in ; The data optimization clustering layer is used to classify the data features T of all patients with chronic respiratory diseases n The optimized kernel fuzzy C-means clustering algorithm with Gaussian radial basis function as kernel function is used for clustering to obtain the initial classification label F n ; The principal component analysis layer of respiratory chronic diseases is used to analyze the data features of all patients with respiratory chronic diseases. n Perform principal component analysis to obtain the main characteristic data Z of patients with chronic respiratory diseases n ; The main feature output layer outputs the main feature data Z of patients with chronic respiratory diseases n And the corresponding initial classification label F n ; The specific steps of clustering in the data optimization clustering layer include: Construct K dynamic cluster center individuals G k , k=1,2,…,K; each dynamic cluster center individual G k Contains the solution of clustering centers for optimizing the kernel fuzzy C-means clustering algorithm; K dynamic clustering center individuals G k Combine to obtain the dynamic clustering iterative population of chronic respiratory diseases; set the maximum number of iterations; Dynamic clustering center individual G in iterative population based on dynamic clustering of chronic respiratory diseases k Data characteristics of patients with chronic respiratory diseases T n Clustering is performed using the optimized kernel fuzzy C-means clustering algorithm with Gaussian radial basis function as the kernel function to obtain simulated clustering results; clustering effect evaluation is performed on the simulated clustering results to obtain the clustering effect evaluation result L k ; The clustering effect evaluation result L k As the dynamic cluster center individual G k The fitness S k ; When the maximum number of iterations is reached, the dynamic clustering center individual G with the largest fitness in the dynamic clustering iteration population of chronic respiratory diseases is output. k , which is the optimal dynamic clustering center individual; the solution of the clustering center in the optimal dynamic clustering center individual is used as the solution of the initial classification label, and the data feature T of all respiratory chronic disease patients after clustering is n Match the initial classification label F n ; The classification model for clinical research on chronic respiratory diseases in step S2 includes a data feature fusion layer, a data feature classification layer and a classification result output layer; The data feature fusion layer is used to transform the main feature data Z of patients with chronic respiratory diseases n and the initial classification label F n Perform feature fusion to obtain the fusion feature R of patients with chronic respiratory diseases n ; The data feature classification layer is used to fuse the features R according to the patients with chronic respiratory diseases. n Perform feature analysis to obtain the classification results of patients with chronic respiratory diseases B n ; The classification result output layer is used to output the classification results of patients with chronic respiratory diseases B n ; The specific steps of training data feature classification layer include: Collect several groups of respiratory chronic disease typing matching training samples; each group of respiratory chronic disease typing matching training samples contains a group of historical respiratory chronic disease patient fusion features and corresponding respiratory chronic disease patient typing results, and the corresponding respiratory chronic disease patient typing results are used as the target value of each group of respiratory chronic disease typing matching training samples; combine several groups of respiratory chronic disease typing matching training samples to obtain a respiratory chronic disease typing matching training set; The respiratory chronic disease typing matching training set is input into the respiratory chronic disease clinical research classification model to train the data feature classification layer with the target value as the target to obtain the initial data feature classification layer; the initial data feature classification layer is model evaluated to obtain the initial data feature classification layer model evaluation result; if the initial data feature classification layer model evaluation result is passed, the trained initial data feature classification layer is used as the data feature classification layer in the respiratory chronic disease clinical research classification model; otherwise, the respiratory chronic disease typing matching training set is used to continue model training.

2. A method for clinical research and management of chronic respiratory diseases according to claim 1, characterized in that: According to the classification results of patients with chronic respiratory diseases B n Match the corresponding initial chronic respiratory disease patient tracking treatment plan A n The specific steps include: Obtain a standard library of tracking and treatment plans for patients with chronic respiratory diseases, which contains several groups of standard chronic respiratory disease patient classification results and corresponding standard chronic respiratory disease patient tracking and treatment plans; The results of the classification of patients with chronic respiratory diseases are B n Match the most similar standard respiratory chronic disease patient classification result in the standard database of respiratory chronic disease patient tracking treatment plans, and use the standard respiratory chronic disease patient tracking treatment plan corresponding to the most similar standard respiratory chronic disease patient classification result as respiratory chronic disease patient classification result B. n Corresponding initial chronic respiratory disease patient follow-up treatment plan A n .

3. A method for clinical research and management of chronic respiratory diseases according to claim 2, characterized in that: The dynamic decision-making model for clinical research on chronic respiratory diseases in step S3 includes a text feature extraction layer, a difference dynamic factor analysis layer, and an optimization solution output layer; The text feature extraction layer is used to extract the dynamic decision dataset V n Extract features from the decision data in the dynamic decision data feature set V n ',V n '={V n1 , V n2 , …, V nj , …, V nJ }, J is the dynamic decision data feature set V n 'The number of decision data item features, V nj Represents the dynamic decision data feature set V n 'The decision data item characteristics in '; The differential dynamic factor analysis layer is used to analyze the dynamic decision data feature set V n 'Perform feature recognition and obtain the decision data difference factor set C n , C n ={C n1 , C n2 , …, C nj , …, C nJ }, C nj Represents the sub-item features of decision data V nj The corresponding matching decision data difference factor; The optimization solution output layer is used to calculate the difference factor set C based on the decision data. n Follow-up treatment plan A for patients with initial chronic respiratory diseases n Adjustments were made to obtain an optimized follow-up treatment plan for patients with chronic respiratory diseases A n '.

4. A method for clinical research and management of chronic respiratory diseases according to claim 3, characterized in that: The specific steps for training the differential dynamic factor analysis layer include: Collecting several groups of dynamic difference factor training samples, each group of dynamic difference factor training samples includes a group of prior dynamic decision data features and corresponding difference dynamic factors; using the corresponding difference dynamic factors as target values ​​for each group of dynamic difference factor training samples; combining several groups of dynamic difference factor training samples to obtain a dynamic difference factor training set; The dynamic differential factor training set is input into the dynamic decision-making model for clinical research on chronic respiratory diseases, and the differential dynamic factor analysis layer is trained with the target value as the target to obtain the initial differential dynamic factor analysis layer; the initial differential dynamic factor analysis layer is evaluated to obtain the model evaluation result of the initial differential dynamic factor analysis layer; if the model evaluation result of the initial differential dynamic factor analysis layer is passed, the trained initial differential dynamic factor analysis layer is used as the differential dynamic factor analysis layer in the dynamic decision-making model for clinical research on chronic respiratory diseases; otherwise, the dynamic differential factor training set is used to continue model training.

Images