Intelligent analysis and prediction platform for curative effect of traditional Chinese medicine

Abstract

The invention discloses a traditional Chinese medicine curative effect intelligent analysis and prediction platform, and relates to the technical field of traditional Chinese medicine analysis. The intelligent management platform is in communication connection with a data preprocessing and standardization module, a symptom description standardization processing module, a network relation analysis module, a network target positioning module, a network target navigation module, a drug effect prediction and evaluation module and a report generation and output module. According to the traditional Chinese medicine curative effect intelligent analysis and prediction platform, by integrating a machine learning technology, the accuracy of traditional Chinese medicine curative effect evaluation is remarkably improved, a large number of traditional Chinese medicine prescriptions, patient syndromes and curative effect data information are collected and analyzed, and the treatment effects of the traditional Chinese medicine prescriptions on different syndromes are evaluated by combining a network target and an established network relationship; and analysis results of the drug target intervention effect and the synergistic effect are provided, so that the action mechanism of the traditional Chinese medicine prescription can be deeply understood, and the traditional Chinese medicine prescription with a potential curative effect can be quickly screened out.

Description

Technical Field

[0001] The present invention relates to the technical field of traditional Chinese medicine analysis, and in particular to an intelligent analysis and prediction platform for the efficacy of traditional Chinese medicine. Background Art

[0002] With the advancement of science and technology and people's deepening understanding of traditional Chinese medicine, the modernization of traditional Chinese medicine has become an inevitable trend. Traditional Chinese medicine, which relies on experience and oral transmission, can no longer meet the needs of modern medical care. In recent years, the rapid development of big data and artificial intelligence technology has provided strong support for the intelligent analysis and prediction of the efficacy of traditional Chinese medicine. By collecting and analyzing a large amount of clinical data on traditional Chinese medicine and using machine learning, deep learning and other algorithms, we can explore the potential laws and influencing factors of the efficacy of traditional Chinese medicine, thereby providing a scientific basis for the prediction of the efficacy of traditional Chinese medicine. Therefore, using modern information technology to conduct intelligent analysis and prediction of the efficacy of traditional Chinese medicine has become an important means to promote the modernization of traditional Chinese medicine.

[0003] Clinical data of traditional Chinese medicine are diverse, complex and non-standardized, resulting in low data quality and difficulty in intelligent analysis and prediction. In addition, there are problems such as diverse symptom descriptions and difficulty in standardization. Therefore, how to standardize symptom descriptions and establish a unified symptom terminology library and efficacy evaluation standards to achieve predictive analysis of the efficacy of traditional Chinese medicine is a problem we need to solve. To this end, this paper proposes an intelligent analysis and prediction platform for the efficacy of traditional Chinese medicine. Summary of the invention

[0004] To achieve the above purpose, the present invention is implemented by the following technical solutions: a Chinese medicine efficacy intelligent analysis and prediction platform, including an intelligent management platform, the intelligent management platform is communicatively connected with a data preprocessing and standardization module, a symptom description standardization processing module, a network relationship analysis module, a network target positioning module, a network target navigation module, a drug efficacy prediction and evaluation module and a report generation and output module, wherein the modules are connected by electrical signals;

[0005] The data preprocessing and standardization module is used to obtain clinical data of traditional Chinese medicine, and preprocess and standardize the clinical data of traditional Chinese medicine to solve the problems of data diversity, complexity and non-standardization, improve the data quality, make it suitable for intelligent analysis and prediction, and provide high-quality input data for subsequent modules;

[0006] The symptom description standardization processing module is used to standardize symptom descriptions, establish a unified symptom terminology library and efficacy evaluation criteria, so as to standardize symptom descriptions, achieve consistency and comparability of symptom descriptions, and facilitate symptom-related data analysis and prediction;

[0007] The network relationship analysis module uses natural language processing technology to conduct in-depth mining and analysis of TCM clinical data, analyze the relationship between syndromes and drug-level objects, provide basic data support for subsequent network target positioning, reveal the complex relationship between syndromes and drugs, and provide a network basis for intelligent analysis of TCM efficacy;

[0008] The network target positioning module predicts the relationship between syndromes and drug targets based on the analysis results at the syndrome and drug levels, identifies the network target nodes of Chinese medicines that affect syndromes, and provides precise targets for drug efficacy prediction;

[0009] The network target navigation module analyzes the specific mechanism of action of the drug through disease-disease relationship analysis, drug-disease relationship analysis, and drug-drug relationship analysis, and establishes a network relationship between the drug and the syndrome;

[0010] The drug efficacy prediction and evaluation module constructs a Chinese medicine efficacy prediction model based on network targets and established network relationships, predicts the efficacy of Chinese medicine, and evaluates it, providing a quantitative prediction of the potential anti-disease efficacy of Chinese medicine prescriptions, and providing a scientific basis for clinical drug use;

[0011] The report generation and output module automatically generates a Chinese medicine efficacy analysis report based on user needs and query results for visual display.

[0012] Preferably, in the data preprocessing and standardization module, the process of preprocessing and standardizing the TCM clinical data includes:

[0013] Collect TCM clinical data from multiple data sources including electronic health records, clinical trials, and research databases. TCM clinical data include patient symptom descriptions, diagnosis results, prescription information, and treatment effects;

[0014] Preprocessing the collected clinical data of traditional Chinese medicine, including data cleaning and data formatting;

[0015] The preprocessed data is standardized. For symptom descriptions and diagnosis results, natural language processing technology is used to perform text analysis, extract keywords and convert them into numerical features. For prescription information, it is encoded according to the type and dosage characteristics of the drug to form a numerical data set.

[0016] Symptoms, diagnosis results and prescription information data are coded and classified according to coding rules. Symptoms are assigned unified codes based on traditional Chinese medicine theory, diagnosis results are coded using the ICD medical classification system, and prescription information is coded and classified according to the ingredients and efficacy characteristics of the drugs to form a unified drug analysis data set.

[0017] Preferably, in the symptom description standardization processing module, the process of establishing the symptom terminology library and the efficacy evaluation criteria includes:

[0018] Extract original symptom descriptions from pre-processed TCM clinical data to ensure that the collected symptom descriptions are comprehensive and detailed, covering various diseases and symptom types, and organize and analyze the collected symptom descriptions to identify and remove repeated, redundant or inconsistent terms;

[0019] Refer to the symptom classification standards published by TCM professional books, literature and authoritative institutions, use multi-label classification method to assign multiple semantic labels to each symptom description, perform semantic division, obtain candidate standard symptom words under the corresponding semantic labels, and use the symptom word matching model to score and sort the candidate symptom word set. Based on the similarity calculation, the candidate word with the highest score under each semantic label is selected as the final standardized result to ensure that the standardized symptom terms have clear and consistent definitions and descriptions;

[0020] Integrate the standardized symptom terms and enter them into the database to establish a unified symptom terminology library. The terminology library includes the definition, description, related diseases and symptom information of each symptom term, ensuring that each symptom term in the terminology library has a clear definition and description to facilitate subsequent data analysis and prediction;

[0021] According to TCM theory and clinical practice, the related indicators for efficacy evaluation are determined, namely, symptom improvement, changes in laboratory test results, and improvement in patients' quality of life, to ensure that the evaluation indicators are scientific, comprehensive, and operational. Each evaluation indicator is refined and quantified, and specific evaluation standards are formulated. For symptom improvement, indicators of symptom improvement rate and symptom disappearance time are formulated. For changes in laboratory test results, indicators with specific numerical ranges are formulated.

[0022] Integrate all evaluation indicators and evaluation standards to establish a complete efficacy evaluation system.

[0023] Preferably, in the network relationship analysis module, the mining and analysis process of TCM clinical data includes:

[0024] Use natural language processing technology to extract clinical symptom information related to syndromes from databases and literature in the field of traditional Chinese medicine, clean, remove duplicates and normalize the extracted information, and build a syndrome-clinical symptom relationship database;

[0025] Use natural language processing technology to extract the properties and relationship information of Chinese herbal medicines from traditional Chinese medicine databases, ancient literature and modern research literature, and organize the extracted information to build a Chinese herbal medicine property relationship database, including the nature, flavor, meridians, efficacy and indications of Chinese herbal medicines;

[0026] Access public databases (such as STRING, Reactome, etc.), download protein-protein interaction, signaling pathway and transcriptional regulation relationship data, and pre-process the downloaded data, including data cleaning and format conversion, for subsequent analysis;

[0027] Based on omics data (gene expression data), co-expression networks were constructed to analyze the co-expression relationships between different genes. The network topology method was used to evaluate the similarity of network modules composed of different phenotypes belonging to the same syndrome or different syndromes of the same disease. The Bayesian network model was used to integrate information from different data sources to construct a syndrome-symptom-drug relationship network. The causal inference method (causal reasoning based on Bayesian network) was applied to analyze the causal relationship between nodes in the network. The network topology method (modularity or clustering coefficient) was used to evaluate the similarity of different network modules. The network modules composed of different herbs in the same prescription were analyzed to evaluate the synergistic effect between herbs within the prescription.

[0028] Combining the results of co-expression network analysis and Bayesian model integration, the relationship between syndromes and drug efficacy is analyzed, drugs or prescriptions with significant therapeutic effects on specific syndromes are identified, and their mechanisms of action are explored. The constructed drug relationship network is used to analyze the interaction relationship between drugs, identify the synergistic or antagonistic effects between drugs, and provide a theoretical basis for the combination of drugs. Combining the Chinese herbal medicine attribute relationship database and protein-protein interaction data, the relationship between herbal targets is integrated, the interaction between different herbal targets is analyzed, and the synergistic mechanism of herbal medicines in prescriptions is explored.

[0029] Preferably, in the network target positioning module, the identification process of the network target node includes:

[0030] By collecting and analyzing syndrome-related clinical symptoms, the properties and relationships of Chinese medicinal herbs, and protein-protein interaction, signaling, and transcriptional regulation relationship data extracted from public databases, a multi-level biological network of syndrome-drug-target was constructed;

[0031] Use natural language processing technology and data mining methods to analyze the relationship between syndromes and drugs, including the effect of drugs on improving syndromes and the role of drugs in syndrome treatment. Based on the interaction data between drugs and targets, mine the association between drugs and targets and identify potential targets for drug action.

[0032] In the constructed multi-level biological network, network topology analysis technology is used to locate the network target nodes that affect syndromes;

[0033] Based on the positioning results of network targets, machine learning models are used to predict the relationship between syndromes and drug targets, and to analyze the intervention effects of drug targets on syndromes and the synergistic effects of drug targets in syndrome treatment.

[0034] Preferably, the process of predicting the relationship between syndromes and drug targets includes:

[0035] Integrate data on syndromes, drugs, and targets, and collect data on the interactions between syndromes, drugs, and targets to form a comprehensive data set;

[0036] Combined with the network target nodes closely related to the syndrome located in the multi-level biological network, the properties of the network target and its position in the network are clarified, the preliminary impact of the network target on the syndrome is analyzed, and the acquired data sets are integrated to extract the correlation features of the network target prediction. Among them, the network target attributes include protein type, functional classification, and connectivity in the biological network, and the features include the clinical manifestations of the syndrome, the chemical properties of the drug, and the amino acid sequence and structure of the target;

[0037] The network target prediction model is constructed by using the characteristic data in the data set combined with machine learning technology, and the model parameters are tuned and trained according to the characteristics of the data set until the model achieves stable prediction performance to improve the accuracy of the prediction;

[0038] Use the trained network target prediction model, input the characteristic data of syndromes and drug targets, and predict the relationship between syndromes and drug targets;

[0039] According to the prediction results and the properties of the network targets, combined with the total intervention probability of the drug target on all symptoms and the ratio of the baseline intervention effect of the drug target to the extreme intervention effect of the drug target, the network target evaluation index was calculated to quantify the intervention effect and synergistic effect of the network target on the syndrome.

[0040] Preferably, in the network target navigation module, the process of establishing the drug-syndrome network relationship includes:

[0041] Collect relevant information on various diseases, including clinical manifestations, diagnostic criteria, and pathophysiological processes of the diseases, and use network science methods to consider diseases as nodes and the relationships between diseases as edges, build disease networks, and analyze the characteristics of nodes and edges in disease networks to identify key nodes and modules in disease networks, and then analyze the relationships between diseases;

[0042] Collect relevant information about drugs, including their chemical structure, target, and pharmacodynamic properties. Combine the target information with the disease network to analyze the specific mechanism of action of drugs on diseases. Consider drugs and diseases as nodes, and the relationship between drugs and diseases (such as therapeutic effects, side effects, etc.) as edges. Build a drug-disease network, analyze the connection between drug nodes and disease nodes in the network, evaluate the efficacy and safety of drugs, and identify network targets closely related to drug effects in the drug-disease network.

[0043] Collect information on interactions between different drugs, including synergistic effects and antagonistic effects, analyze the mechanism of drug interactions, including whether they are through common targets, signaling pathways or metabolic pathways, consider drugs as nodes, and the interaction relationships between drugs as edges, build a drug network, analyze the characteristics of nodes and edges in the drug network, identify the associations and similarities between drugs, optimize drug combination plans based on the analysis results of the drug network, and improve the therapeutic effect and reduce side effects by combining drugs with synergistic effects;

[0044] Collect relevant information on TCM syndromes, including the clinical manifestations, diagnostic criteria, etiology and pathogenesis of the syndromes, integrate syndrome data from different sources, regard drugs and syndromes as nodes, and the relationship between drugs and syndromes as edges, construct a drug-syndrome network, analyze the connection between drug nodes and syndrome nodes in the network, evaluate the efficacy and indication range of drugs on syndromes, combine the identification results of network targets and the analysis results of the drug-syndrome network, and identify the association between network targets and syndromes.

[0045] Preferably, in the efficacy prediction and evaluation module, the prediction process of the efficacy of traditional Chinese medicine includes:

[0046] Based on the analysis results of disease network, drug-disease network and drug network, a multi-level biological network of syndrome-drug-target is constructed;

[0047] Collect a large number of Chinese herbal prescriptions, their corresponding target information and related network relationship data, and use reverse molecular docking technology to dock the active ingredients in the Chinese herbal prescriptions with the known target database. Through spatial matching and energy calculation, screen out the targets of the Chinese herbal prescriptions, establish the interaction relationship between the Chinese herbal prescriptions and the network targets, and form a network model;

[0048] Extract syndrome characteristics, drug characteristics, and target characteristics from multi-level biological networks, use machine learning technology to build a TCM efficacy prediction model, and optimize the model to improve prediction accuracy and stability, and use the TCM efficacy prediction model to predict the efficacy of TCM;

[0049] The efficacy prediction model of traditional Chinese medicine is combined with network targets and established network relationships to predict the efficacy of traditional Chinese medicine prescriptions. Based on the prediction results, the efficacy evaluation index is calculated to analyze the intervention effect of drug targets on syndromes and the synergistic effect of drug targets in syndrome treatment.

[0050] The Chinese medicine prescriptions are evaluated based on the analysis results of the efficacy evaluation index and drug targets. Based on the evaluation results, the Chinese medicine prescriptions are optimized, including adjusting the drug composition and dosage, to improve efficacy and safety.

[0051] Preferably, the calculation expression of the efficacy evaluation index is:

[0052] ;

[0053] in, is the efficacy evaluation index, For drug The intervention effect of the network target is the baseline value, which represents the average efficacy of the drug without intervention. The maximum intervention effect of the drug is obtained from literature or databases. It is the range limit of the drug intervention effect, reflecting the possible range of drug intervention effects. is the total number of nodes in the drug network, representing the number of targets that the drug may affect. is the number of nodes directly connected to the drug in the network, representing the number of targets directly affected by the drug. is a constant used to adjust the sensitivity of the model.

[0054] Preferably, in the report generation and output module, the generation process of the Chinese medicine efficacy analysis report includes:

[0055] Summarize the output results of the efficacy prediction and evaluation module, including the efficacy evaluation index, network target analysis results, and drug-syndrome network analysis, and collect relevant information of the user's query, parse the query conditions entered by the user, determine the database and query statement to be queried, and clarify the user's query conditions;

[0056] According to the parsed query conditions, query the database for relevant Chinese medicine prescriptions, syndromes, efficacy evaluation, efficacy evaluation index, and analyze the intervention effect and synergistic effect data of drug targets;

[0057] Fill the query and processed data into the report template, write the summary and conclusion of the report, summarize the efficacy and mechanism of Chinese medicine prescriptions on specific syndromes, and edit and proofread the generated Chinese medicine efficacy analysis report to ensure accurate language, clear logic, and no grammatical errors;

[0058] The generated TCM efficacy analysis report is output in a user-readable format and displayed on the intelligent management platform for users to review and share.

[0059] The present invention provides a platform for intelligent analysis and prediction of the efficacy of traditional Chinese medicines. It has the following beneficial effects:

[0060] 1. This intelligent analysis and prediction platform for the efficacy of traditional Chinese medicine significantly improves the accuracy of the evaluation of the efficacy of traditional Chinese medicine by integrating machine learning technology. It collects and analyzes a large amount of traditional Chinese medicine prescriptions, patient symptoms, and efficacy data information. It combines network targets with established network relationships to evaluate the therapeutic effects of traditional Chinese medicine prescriptions on different symptoms, and provides analysis results of drug target intervention and synergistic effects, which helps to deeply understand the mechanism of action of traditional Chinese medicine prescriptions, and then quickly screen out traditional Chinese medicine prescriptions with potential efficacy.

[0061] 2. This intelligent analysis and prediction platform for the efficacy of traditional Chinese medicine has significantly improved the accuracy and efficiency of the analysis of the efficacy of traditional Chinese medicine by integrating multiple data sources. It collects and analyzes syndrome-related clinical symptoms, the properties and relationship data of medicinal and herbal medicines, and the protein-protein interaction, signal and transcriptional regulation relationship data extracted from public databases, providing rich information for the construction of disease networks, drug-disease networks and drug networks. By using natural language processing technology and data mining methods, it can deeply mine and analyze these data, reveal the complex relationship between syndromes and drugs, and provide a network foundation for the intelligent analysis of the efficacy of traditional Chinese medicine.

[0062] 3. The intelligent analysis and prediction platform for the efficacy of traditional Chinese medicine predicts the relationship between syndromes and drug targets through machine learning models, calculates the efficacy evaluation index, predicts the efficacy of traditional Chinese medicine, and conducts evaluation. This not only improves the accuracy of the analysis, but also greatly shortens the analysis time, providing a rapid scientific basis for clinical drug use. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 This is a module structure diagram of a traditional Chinese medicine efficacy intelligent analysis and prediction platform of the present invention;

[0064] Figure 2 The present invention is a flowchart of mining and analyzing the clinical data of traditional Chinese medicine;

[0065] Figure 3 A flow chart for predicting the relationship between the syndrome and the drug target of the present invention;

[0066] Figure 4 It is a flow chart for predicting the efficacy of the Chinese medicine of the present invention. DETAILED DESCRIPTION

[0067] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. The embodiments of the present invention are provided for the purpose of illustration and description, and are not intended to be exhaustive or to limit the present invention to the disclosed forms. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments are selected and described in order to better illustrate the principles and practical applications of the present invention, and to enable those of ordinary skill in the art to understand the present invention and thereby design various embodiments with various modifications suitable for specific uses.

[0068] The first embodiment, as Figures 1 to 3 As shown, the present invention provides a technical solution: a Chinese medicine efficacy intelligent analysis and prediction platform, including an intelligent management platform, the intelligent management platform is communicatively connected with a data preprocessing and standardization module, a symptom description standardization processing module, a network relationship analysis module, a network target positioning module, a network target navigation module, a drug efficacy prediction and evaluation module, and a report generation and output module, wherein the modules are connected by electrical signals;

[0069] The data preprocessing and standardization module is used to obtain clinical data of traditional Chinese medicine, and to preprocess and standardize the clinical data of traditional Chinese medicine to solve the problems of data diversity, complexity and non-standardization, improve data quality, make it suitable for intelligent analysis and prediction, and provide high-quality input data for subsequent modules. Clinical data of traditional Chinese medicine are collected from multiple data sources such as electronic health records, clinical trials, and research databases. Clinical data of traditional Chinese medicine include patient symptom descriptions, diagnosis results, prescription information, and treatment effects. The collected clinical data of traditional Chinese medicine are preprocessed. The preprocessing operation includes data cleaning and data formatting steps. Among them, data cleaning is used to identify and correct errors and inconsistencies in the data, such as removing duplicate records, filling missing values, and correcting obvious data entry errors to ensure the accuracy of the data. Data formatting is used to convert data into a unified format, such as the standardization of the formats of date, time, numerical and categorical data, so that the data is consistent in structure and easy to process and analyze. The pre-processed data is standardized. For symptom descriptions and diagnostic results, natural language processing technology is used for text analysis, keywords are extracted and converted into numerical features. For prescription information, it is encoded according to the type and dosage characteristics of the drug to form a numerical data set. Symptoms, diagnostic results and prescription information data are encoded and classified according to coding rules. Symptoms are assigned a unified code according to traditional Chinese medicine theory. Diagnostic results are encoded using the ICD medical classification system. Prescription information is encoded and classified according to the composition and efficacy characteristics of the drug to form a unified drug analysis data set.

[0070] The symptom description standardization processing module is used to standardize symptom descriptions, establish a unified symptom terminology library and efficacy evaluation standards to standardize symptom expressions, achieve consistency and comparability of symptom descriptions, and facilitate symptom-related data analysis and prediction. The original symptom descriptions are extracted from the pre-processed TCM clinical data to ensure that the collected symptom descriptions are comprehensive and detailed, covering various diseases and symptom types. The collected symptom descriptions are sorted and analyzed, and repeated, redundant or inconsistent terms are identified and removed. Referring to the symptom classification standards published by TCM professional books, literature and authoritative organizations, a multi-label classification method is used to assign multiple semantic labels to each symptom description, perform semantic division, and obtain candidate standard symptom words under the corresponding semantic labels. The symptom word matching model is used to score and sort the candidate symptom word set, and the candidate word with the highest score under each semantic label is selected as the final standardization result based on similarity calculation to ensure that the standardized symptom terminology is clear and consistent. The definition and description of the symptom terms after normalization are integrated and entered into the database to establish a unified symptom terminology library. The terminology library includes the definition, description, related diseases and symptom information of each symptom term, ensuring that each symptom term in the terminology library has a clear definition and description, which is convenient for subsequent data analysis and prediction. According to the theory of traditional Chinese medicine and clinical practice, the related indicators for efficacy evaluation are determined, which are symptom improvement, changes in laboratory test results, and improvement in the quality of life of patients, ensuring that the evaluation indicators are scientific, comprehensive and operational, and each evaluation indicator is refined and quantified, and specific evaluation standards are formulated. For symptom improvement, indicators of symptom improvement rate and symptom disappearance time are formulated. For changes in laboratory test results, indicators of specific numerical ranges are formulated. All evaluation indicators and evaluation standards are integrated to establish a complete efficacy evaluation system, ensuring that the evaluation system can comprehensively and objectively reflect the efficacy of traditional Chinese medicine treatment, and provide strong support for subsequent data analysis and prediction;

[0071] The network relationship analysis module uses natural language processing technology to conduct in-depth mining and analysis of TCM clinical data, analyze the relationship between syndrome and drug-level objects, provide basic data support for subsequent network target positioning, reveal the complex relationship between syndrome and drug, and provide a network foundation for intelligent analysis of the efficacy of traditional Chinese medicine. It uses natural language processing technology to extract clinical symptom information related to syndromes from databases and literature in the field of traditional Chinese medicine, cleans, removes duplicates and normalizes the extracted information, and builds a syndrome-clinical symptom relationship database. It uses natural language processing technology to extract the properties and relationship information of Chinese herbal medicines from traditional Chinese medicine databases, ancient literature and modern research literature, and organizes the extracted information to build a Chinese herbal medicine property relationship database, including the nature, flavor, meridians, and efficacy of Chinese herbal medicines. It accesses public databases (such as STRING, Reactome, etc.), downloads protein-protein interaction, signal pathway and transcriptional regulation relationship data, and pre-processes the downloaded data, including data cleaning and format conversion, for subsequent analysis. It builds a co-expression network based on omics data (gene expression data) and analyzes the relationship between different genes. co-expression relationship, and use network topology methods to evaluate the similarity of network modules composed of different phenotypes belonging to the same syndrome or different syndromes of the same disease, use Bayesian network models to integrate information from different data sources, construct syndrome-symptom-drug relationship networks, apply causal inference methods (causal reasoning based on Bayesian networks) to analyze the causal relationship between nodes in the network, use network topology methods (modularity or clustering coefficient) to evaluate the similarity of different network modules, analyze the similarity of network modules composed of different herbs in the same prescription to evaluate the synergistic effect between herbs within the prescription, combine the results of co-expression network analysis and Bayesian model integration, analyze the relationship between syndromes and drug efficacy, identify drugs or prescriptions with significant efficacy for specific syndromes, and explore their mechanism of action, use the constructed drug relationship network to analyze the interaction relationship between drugs, identify synergistic or antagonistic effects between drugs, and provide a theoretical basis for the combined use of drugs, combine the Chinese herbal medicine attribute relationship database and protein-protein interaction data, integrate the relationship between herbal targets, analyze the interaction between different herbal targets, and explore the synergistic mechanism of herbal drugs in prescriptions;

[0072] The network target positioning module predicts the relationship between syndromes and drug targets based on the analysis results at the syndrome and drug levels, identifies the network target nodes of Chinese medicine that affect syndromes, and provides precise targets for drug efficacy prediction. By collecting and analyzing syndrome-related clinical symptoms, the properties and relationships of Chinese herbal medicines, and protein-protein interactions, signals, and transcriptional regulation relationship data extracted from public databases, a multi-level biological network of syndromes, drugs, and targets is constructed. The nodes in the multi-level biological network represent syndromes, drugs, proteins, and genetic biomolecules, and the edges represent the interactions between nodes. Natural language processing technology and data mining methods are used to analyze syndromes. The relationship between drugs and syndromes, including the effect of drugs on improving syndromes and the status of drugs in syndrome treatment. Based on the interaction data between drugs and targets, the association between drug targets is mined, and potential targets of drug action are identified. In the constructed multi-level biological network, network target nodes that affect syndromes are located using network topology analysis technology. Network target nodes are proteins, genes or other biological molecules that play an important role in the occurrence and development of syndromes. Based on the positioning results of network targets, machine learning models are used to predict the relationship between syndromes and drug targets, and the intervention effect of drug targets on syndromes and the synergistic effect of drug targets in syndrome treatment are analyzed.

[0073] The process of predicting the relationship between syndromes and drug targets includes:

[0074] Integrate the data of syndromes, drugs, and targets, and collect the interaction relationship data between syndromes, drugs, and targets to form a comprehensive data set. Combine the network target nodes closely related to the syndrome located in the multi-level biological network, clarify the properties of the network target and its position in the network, analyze the initial impact of the network target on the syndrome, and comprehensively obtain the acquired data set to extract the correlation features of the network target prediction. Among them, the network target attributes include protein type, functional classification, and connectivity in the biological network. The features include the clinical manifestations of the syndrome, the chemical properties of the drug, and the amino acid sequence and structure of the target. The characteristic data is combined with machine learning technology to construct a network target prediction model, and according to the characteristics of the data set, the model parameters are tuned and trained until the model reaches a stable prediction performance to improve the accuracy of the prediction. The trained network target prediction model is used to input the characteristic data of syndromes and drug targets to predict the relationship between syndromes and drug targets. According to the prediction results and the properties of the network target, combined with the total intervention probability of the drug target on all symptoms and the ratio of the baseline intervention effect of the drug target to the extreme intervention effect of the drug target, the network target evaluation index is calculated to quantify the intervention effect and synergistic effect of the network target on the syndrome;

[0075] Furthermore, the calculation expression of the network target evaluation index is:

[0076] ;

[0077] in, is the network target evaluation index, indicating the drug target Symptoms The probability of intervention is the set of symptoms related to the syndrome, As drug targets, For symptoms, Drug target The baseline intervention effect Drug target The ultimate intervention effect means that under ideal conditions The maximum intervention effect that can be achieved Drug target Total intervention probability for all symptoms, is the ratio of the baseline value to the limit value, which is used to adjust the intervention probability. The value range of is limited to between 0 and 1, where 0 means that the drug target has no intervention effect on the syndrome, and 1 means that the drug target achieves its maximum possible intervention effect. The higher the value, the stronger the intervention effect of the drug target on the syndrome and the more obvious the synergistic effect;

[0078] The network target navigation module analyzes the specific mechanism of action of drugs through disease-disease relationship analysis, drug-disease relationship analysis, and drug-drug relationship analysis, and establishes a network relationship between drugs and syndromes, collects relevant information of various diseases, including the clinical manifestations, diagnostic criteria, and pathophysiological processes of the diseases, and uses network science methods to regard diseases as nodes and the relationships between diseases as edges to construct disease networks, analyze the characteristics of nodes and edges in disease networks to identify key nodes and modules in disease networks, and then analyze the relationships between diseases, collect relevant information of drugs, including the chemical structure, target, and pharmacodynamic properties of drugs, combine drug target information and disease networks to parse the specific mechanism of action of drugs on diseases, regard drugs and diseases as nodes, and the relationships between drugs and diseases (such as therapeutic effects, side effects, etc.) as edges to construct drug-disease networks, analyze the connection between drug nodes and disease nodes in the network, evaluate the efficacy and safety of drugs, and identify key nodes and modules in drug-disease networks that are related to drug action. Use closely related network targets to collect information on interactions between different drugs, including synergistic effects and antagonistic effects, analyze the mechanism of drug interactions, including whether they are through common targets, signal pathways or metabolic pathways, regard drugs as nodes, and the interaction relationships between drugs as edges, build a drug network, analyze the characteristics of nodes and edges in the drug network, identify the associations and similarities between drugs, optimize drug combination plans based on the analysis results of the drug network, improve the treatment effect and reduce side effects by combining drugs with synergistic effects, collect relevant information on TCM syndromes, including the clinical manifestations, diagnostic criteria, etiology and pathogenesis of syndromes, integrate syndrome data from different sources, regard drugs and syndromes as nodes, and the relationship between drugs and syndromes as edges, build a drug-syndrome network, analyze the connection between drug nodes and syndrome nodes in the network, evaluate the efficacy and indication range of drugs on syndromes, and identify the association between network targets and syndromes based on the identification results of network targets and the analysis results of drug-syndrome networks;

[0079] The drug efficacy prediction and evaluation module builds a Chinese medicine efficacy prediction model based on network targets and established network relationships, predicts the efficacy of Chinese medicine, and evaluates it, providing quantitative predictions of the potential anti-disease efficacy of Chinese medicine prescriptions and providing a scientific basis for clinical drug use;

[0080] The report generation and output module automatically generates a Chinese medicine efficacy analysis report based on user needs and query results, and displays it visually.

[0081] The second embodiment is based on the first embodiment. Figure 4 As shown in the figure, in the efficacy prediction and evaluation module, the prediction process of the efficacy of traditional Chinese medicine includes:

[0082] Based on the analysis results of disease network, drug-disease network and drug network, a multi-level biological network of syndrome-drug-target is constructed, a large number of Chinese medicine prescriptions, their corresponding target information and related network relationship data are collected, and the active ingredients in the Chinese medicine prescriptions are docked with the known target database using reverse molecular docking technology. Through spatial matching and energy calculation, the targets of the Chinese medicine prescriptions are screened out, and the interaction relationship between the Chinese medicine prescriptions and the network targets is established to form a network model. The syndrome characteristics, drug characteristics and target characteristics are extracted from the multi-level biological network, and the Chinese medicine prescriptions are constructed using machine learning technology. The drug efficacy prediction model is used to optimize the model to improve the prediction accuracy and stability. The efficacy prediction model of traditional Chinese medicine is used to predict the efficacy of traditional Chinese medicine. The traditional Chinese medicine efficacy prediction model is used to combine network targets and established network relationships to predict the efficacy of traditional Chinese medicine prescriptions. Based on the prediction results, the efficacy evaluation index is calculated, and the intervention effect of drug targets on syndromes and the synergistic effect of drug targets in syndrome treatment are analyzed. Based on the analysis results of the efficacy evaluation index and drug targets, the traditional Chinese medicine prescriptions are evaluated. Based on the evaluation results, the traditional Chinese medicine prescriptions are optimized, including adjusting the drug composition, dosage, etc., to improve the efficacy and safety;

[0083] Furthermore, the calculation expression of the efficacy evaluation index is:

[0084] ;

[0085] in, is the efficacy evaluation index, For drug The intervention effect of the network target is the baseline value, which represents the average efficacy of the drug without intervention. The maximum intervention effect of the drug is obtained from literature or databases. It is the range limit of the drug intervention effect, reflecting the possible range of drug intervention effects. is the total number of nodes in the drug network, representing the number of targets that the drug may affect. is the number of nodes directly connected to the drug in the network, representing the number of targets directly affected by the drug. is a constant used to adjust the sensitivity of the model. is the sum of the intervention effects of all network targets, reflecting the impact of the drug on the entire network. It is the attenuation factor for the drug to achieve the maximum intervention effect, ensuring that when the drug intervention effect approaches the maximum value, the index will not increase indefinitely. is a logistic function that reflects the effect of the drug on the remaining targets in the network, The value range of is limited to between 0 and 1, where 0 means that the drug has no effect and 1 means that the drug has a perfect effect. The increase and of reduction, The value will increase, indicating that the drug's intervention effect on the network target is enhanced and the efficacy is improved. On the contrary, if Increase, that is, the number of targets directly affected by the drug decreases, The value will decrease, indicating a decrease in efficacy;

[0086] In the report generation and output module, the generation process of the TCM efficacy analysis report includes:

[0087] Summarize the output results of the efficacy prediction and evaluation module, including the efficacy evaluation index, network target analysis results, and drug-syndrome network analysis, and collect relevant information of user queries, parse the query conditions entered by the user, determine the database and query statements to be queried, clarify the user's query conditions, and query the relevant Chinese medicine prescriptions, syndromes, efficacy evaluation, efficacy evaluation index, and analyze the intervention and synergistic data of drug targets from the database according to the parsed query conditions, fill the queried and processed data into the report template, write the summary and conclusion part of the report, summarize the efficacy and mechanism of Chinese medicine prescriptions on specific syndromes, and edit and proofread the generated Chinese medicine efficacy analysis report to ensure accurate language, clear logic, and no grammatical errors, output the generated Chinese medicine efficacy analysis report in a user-readable format, and display it on the intelligent management platform for users to review and share.

[0088] Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field and related fields without creative work should fall within the scope of protection of the present invention. The structures, devices and operating methods not specifically described and explained in the present invention are implemented according to the conventional means in the field unless otherwise specified and limited.

Claims

1. A traditional Chinese medicine efficacy intelligent analysis and prediction platform, including an intelligent management platform, characterized in that: The intelligent management platform is communicatively connected with a data preprocessing and standardization module, a symptom description standardization processing module, a network relationship analysis module, a network target positioning module, a network target navigation module, a drug efficacy prediction and evaluation module, and a report generation and output module, wherein the modules are connected by electrical signals; The data preprocessing and standardization module is used to obtain clinical data of traditional Chinese medicine, and to preprocess and standardize the clinical data of traditional Chinese medicine; The symptom description standardization processing module is used to standardize the symptom description, establish a unified symptom terminology library and efficacy evaluation criteria, so as to standardize the symptom description; The network relationship analysis module uses natural language processing technology to conduct in-depth mining and analysis of TCM clinical data and analyze the relationship between syndrome and drug level objects; The network target positioning module predicts the relationship between syndromes and drug targets based on the analysis results at the syndrome and drug levels, and identifies the network target nodes of Chinese medicines that affect syndromes; The network target navigation module analyzes the specific mechanism of action of the drug through disease-disease relationship analysis, drug-disease relationship analysis, and drug-drug relationship analysis, and establishes a network relationship between the drug and the syndrome; The drug efficacy prediction and evaluation module constructs a Chinese medicine efficacy prediction model based on network targets and established network relationships, predicts the efficacy of Chinese medicine, and evaluates it; The report generation and output module automatically generates a Chinese medicine efficacy analysis report based on user needs and query results for visual display.

2. The intelligent analysis and prediction platform for the efficacy of traditional Chinese medicine according to claim 1, characterized in that: In the data preprocessing and standardization module, the process of preprocessing and standardization of TCM clinical data includes: Collect TCM clinical data from multiple data sources including electronic health records, clinical trials, and research databases. TCM clinical data include patient symptom descriptions, diagnosis results, prescription information, and treatment effects; Preprocessing the collected clinical data of traditional Chinese medicine, including data cleaning and data formatting; The preprocessed data is standardized. For symptom descriptions and diagnosis results, natural language processing technology is used to perform text analysis, extract keywords and convert them into numerical features. For prescription information, it is encoded according to the type and dosage characteristics of the drug to form a numerical data set. Symptoms, diagnosis results and prescription information data are coded and classified according to coding rules. Symptoms are assigned unified codes based on traditional Chinese medicine theory, diagnosis results are coded using the ICD medical classification system, and prescription information is coded and classified according to the ingredients and efficacy characteristics of the drugs to form a unified drug analysis data set.

3. The intelligent analysis and prediction platform for the efficacy of traditional Chinese medicine according to claim 2, characterized in that: In the symptom description standardization processing module, the process of establishing the symptom terminology library and the efficacy evaluation criteria includes: Extract original symptom descriptions from pre-processed TCM clinical data, and organize and analyze the collected symptom descriptions; Referring to the TCM symptom classification standard, a multi-label classification method is used to assign multiple semantic labels to each symptom description, perform semantic division, obtain candidate standard symptom words under the corresponding semantic labels, and use the symptom word matching model to score and sort the candidate symptom word set. Based on the similarity calculation, the candidate word with the highest score under each semantic label is selected as the final standardized result. Integrate the standardized symptom terms and enter them into the database to establish a unified symptom terminology library, which includes the definition, description, related diseases and symptom information of each symptom term; According to TCM theory and clinical practice, we determined the related indicators for efficacy evaluation, namely, symptom improvement, changes in laboratory test results, and improvement in patients' quality of life. We also refined and quantified each evaluation indicator and formulated specific evaluation standards. Integrate all evaluation indicators and evaluation standards to establish a complete efficacy evaluation system.

4. The intelligent analysis and prediction platform for the efficacy of traditional Chinese medicine according to claim 3, characterized in that: In the network relationship analysis module, the mining and analysis process of TCM clinical data includes: Use natural language processing technology to extract clinical symptom information related to syndromes from databases and literature in the field of traditional Chinese medicine, clean, remove duplicates and normalize the extracted information, and build a syndrome-clinical symptom relationship database; Use natural language processing technology to extract the properties and relationship information of Chinese herbal medicines from traditional Chinese medicine databases, ancient literature and modern research literature, organize the extracted information, and build a Chinese herbal medicine property relationship database; Access public databases, download protein-protein interaction, signaling pathway and transcriptional regulation relationship data, and pre-process the downloaded data; Based on omics data, we constructed a co-expression network, analyzed the co-expression relationship between different genes, and used network topology methods to evaluate the similarity of network modules composed of different phenotypes belonging to the same syndrome or different syndromes of the same disease. We used the Bayesian network model to integrate information from different data sources, constructed a syndrome-symptom-drug relationship network, applied causal inference methods to analyze the causal relationship between nodes in the network, and used network topology methods to evaluate the similarity of different network modules. We also analyzed the similarity of network modules composed of different herbs in the same prescription to evaluate the synergistic effect between herbs within the prescription. Combining the results of co-expression network analysis and Bayesian model integration, the relationship between syndromes and drug efficacy is analyzed, drugs or prescriptions with significant therapeutic effects on specific syndromes are identified, and their mechanisms of action are explored. The constructed drug relationship network is used to analyze the interaction relationship between drugs, identify the synergistic or antagonistic effects between drugs, and combine the Chinese herbal medicine attribute relationship database and protein-protein interaction data to integrate the relationship between herbal targets and analyze the interaction between different herbal targets.

5. The intelligent analysis and prediction platform for the efficacy of traditional Chinese medicine according to claim 4, characterized in that: In the network target positioning module, the identification process of the network target node includes: By collecting and analyzing syndrome-related clinical symptoms, the properties and relationships of Chinese medicinal herbs, and protein-protein interaction, signaling, and transcriptional regulation relationship data extracted from public databases, a multi-level biological network of syndrome-drug-target was constructed; Use natural language processing technology and data mining methods to analyze the relationship between syndromes and drugs, including the effect of drugs on improving syndromes and the role of drugs in syndrome treatment. Based on the interaction data between drugs and targets, mine the association between drugs and targets and identify potential targets for drug action. In the constructed multi-level biological network, network topology analysis technology is used to locate the network target nodes that affect syndromes; Based on the positioning results of network targets, machine learning models are used to predict the relationship between syndromes and drug targets, and to analyze the intervention effects of drug targets on syndromes and the synergistic effects of drug targets in syndrome treatment.

6. The intelligent analysis and prediction platform for the efficacy of traditional Chinese medicine according to claim 5, characterized in that: The prediction process of the relationship between the syndrome and the drug target includes: Integrate data on syndromes, drugs, and targets, and collect data on the interactions between syndromes, drugs, and targets to form a comprehensive data set; Combined with the network target nodes closely related to the syndrome located in the multi-level biological network, the properties of the network target and its position in the network are clarified, the preliminary impact of the network target on the syndrome is analyzed, and the acquired data sets are integrated to extract the correlation features of the network target prediction. Among them, the network target attributes include protein type, functional classification, and connectivity in the biological network, and the features include the clinical manifestations of the syndrome, the chemical properties of the drug, and the amino acid sequence and structure of the target; Use the characteristic data in the data set combined with machine learning technology to build a network target prediction model, and optimize and train the model parameters according to the characteristics of the data set; Use the trained network target prediction model, input the characteristic data of syndromes and drug targets, and predict the relationship between syndromes and drug targets; According to the prediction results and the properties of the network targets, combined with the total intervention probability of the drug target on all symptoms and the ratio of the baseline intervention effect of the drug target to the extreme intervention effect of the drug target, the network target evaluation index was calculated to quantify the intervention effect and synergistic effect of the network target on the syndrome.

7. The intelligent analysis and prediction platform for the efficacy of traditional Chinese medicine according to claim 6, characterized in that: In the network target navigation module, the process of establishing the drug-syndrome network relationship includes: Collect relevant information on various diseases, including clinical manifestations, diagnostic criteria, and pathophysiological processes of the diseases, and use network science methods to consider diseases as nodes and the relationships between diseases as edges, construct disease networks, and analyze the characteristics of nodes and edges in disease networks to identify key nodes in disease networks and further analyze the relationships between diseases; Collect relevant information about drugs, including their chemical structure, target, and pharmacodynamic properties. Combine the target information with the disease network to analyze the specific mechanism of action of drugs on diseases. Consider drugs and diseases as nodes, and the relationship between drugs and diseases as edges. Build a drug-disease network, analyze the connection between drug nodes and disease nodes in the network, evaluate the efficacy and safety of drugs, and identify network targets closely related to drug action in the drug-disease network. Collect information on interactions between different drugs, including synergy and antagonism, analyze the mechanism of drug interactions, regard drugs as nodes and the interactions between drugs as edges, construct a drug network, analyze the characteristics of nodes and edges in the drug network, identify the correlation and similarity between drugs, and optimize drug combination plans based on the analysis results of the drug network; Collect relevant information on TCM syndromes, including the clinical manifestations, diagnostic criteria, etiology and pathogenesis of the syndromes, integrate syndrome data from different sources, regard drugs and syndromes as nodes, and the relationship between drugs and syndromes as edges, construct a drug-syndrome network, analyze the connection between drug nodes and syndrome nodes in the network, evaluate the efficacy and indication range of drugs on syndromes, combine the identification results of network targets and the analysis results of the drug-syndrome network, and identify the association between network targets and syndromes.

8. The intelligent analysis and prediction platform for the efficacy of traditional Chinese medicine according to claim 7, characterized in that: In the efficacy prediction and evaluation module, the prediction process of the efficacy of traditional Chinese medicine includes: Based on the analysis results of disease network, drug-disease network and drug network, a multi-level biological network of syndrome-drug-target is constructed; Collect Chinese herbal prescriptions, their corresponding target information and related network relationship data, and use reverse molecular docking technology to dock the active ingredients in the Chinese herbal prescriptions with the known target database. Through spatial matching and energy calculation, screen out the targets of the Chinese herbal prescriptions, establish the interaction relationship between the Chinese herbal prescriptions and the network targets, and form a network model; Extract syndrome characteristics, drug characteristics, and target characteristics from multi-level biological networks, use machine learning technology to build a TCM efficacy prediction model, optimize the model, and use the TCM efficacy prediction model to predict the efficacy of TCM; The efficacy prediction model of traditional Chinese medicine is combined with network targets and established network relationships to predict the efficacy of traditional Chinese medicine prescriptions. Based on the prediction results, the efficacy evaluation index is calculated to analyze the intervention effect of drug targets on syndromes and the synergistic effect of drug targets in syndrome treatment. The traditional Chinese medicine prescriptions are evaluated based on the analysis results of the efficacy evaluation index and drug targets, and are optimized based on the evaluation results.

9. The intelligent analysis and prediction platform for the efficacy of traditional Chinese medicine according to claim 8, characterized in that: The calculation expression of the efficacy evaluation index is: ； in, is the efficacy evaluation index, For drug The intervention effect of the network target is the baseline value, which represents the average efficacy of the drug without intervention. For the maximum intervention effect of the drug, The range limit of drug intervention effect, is the total number of nodes in the drug network, is the number of nodes directly connected to the drug in the network, is a constant, The value range is limited to between 0 and 1, where 0 means that the drug has no therapeutic effect and 1 means that the drug has a perfect therapeutic effect.

10. The intelligent analysis and prediction platform for efficacy of traditional Chinese medicine according to claim 9, characterized in that: In the report generation and output module, the generation process of the Chinese medicine efficacy analysis report includes: Summarize the output results of the efficacy prediction and evaluation module, including the efficacy evaluation index, network target analysis results, and drug-syndrome network analysis, and collect relevant information of the user's query, parse the query conditions entered by the user, determine the database and query statement to be queried, and clarify the user's query conditions; According to the parsed query conditions, query the database for relevant Chinese medicine prescriptions, syndromes, efficacy evaluation, efficacy evaluation index, and analyze the intervention effect and synergistic effect data of drug targets; Fill the query and processed data into the report template, write the summary and conclusion of the report, summarize the efficacy and mechanism of Chinese medicine prescriptions on specific syndromes, and edit and proofread the generated Chinese medicine efficacy analysis report; The generated TCM efficacy analysis report is output in a user-readable format and displayed on the intelligent management platform for users to review and share.

Images