An intelligent processing system and method for drug comparison data in cardiac rehabilitation process

Abstract

The present invention relates to the technical field of data processing, and provides an intelligent processing system and method for drug comparison data in a cardiac rehabilitation process, including: Chinese medicine data including drug formula data and drug efficacy data, extracting core drug data corresponding to the drug formula data, and mining drug active ingredients and pharmacological information in the core drug data; a pharmacological analysis module, used for classifying the pharmacological information, obtaining the classified pharmacological information, and analyzing the pharmacological similarity between drug active ingredients; a drug efficacy analysis module, used for analyzing the drug efficacy correlation between drug efficacy data; a drug interaction analysis module, used for identifying the treatment cycle corresponding to the Chinese medicine data in the cardiac rehabilitation process, and analyzing the drug interaction between the target drugs in the cycle; and a result generation module, used for generating drug comparison results in the cardiac rehabilitation process. The present invention aims to improve the accuracy of intelligent processing of drug comparison data in the cardiac rehabilitation process.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an intelligent processing system and method for drug comparison data in a cardiac rehabilitation process. Background Art

[0002] Cardiac rehabilitation refers to a series of medical and non-medical interventions to help heart disease patients restore their physical functions, improve their quality of life, and reduce the risk of cardiac events and death. It is a comprehensive, individualized treatment plan that aims to help patients gradually recover to the best state of cardiac function through exercise training, psychological support, nutritional guidance and other means. A large amount of comparative data on traditional Chinese medicines will be collected during the process of cardiac rehabilitation. In order to improve the efficiency of subsequent cardiac rehabilitation, it is necessary to analyze the comparative data of traditional Chinese medicines in order to analyze the efficacy and side effects of different drugs in the process of cardiac rehabilitation, so as to provide more accurate treatment plans and guidance.

[0003] The existing intelligent processing system for drug comparison data in the cardiac rehabilitation process is a decision support system that uses an artificial intelligence algorithm. The system analyzes a large amount of clinical data and research results, extracts information about cardiac rehabilitation treatment, and then provides doctors with guiding treatment suggestions to facilitate the formulation of corresponding treatment plans. However, the system is usually based on existing medical knowledge and research results, which may lead to deviations and limitations in the analysis results, and thus lead to low accuracy in the formulation of subsequent treatment plans. Therefore, there is a need for a system that can improve the accuracy of intelligent processing of drug comparison data in the cardiac rehabilitation process. Summary of the invention

[0004] The present invention provides an intelligent processing system and method for drug comparison data in a cardiac rehabilitation process, the main purpose of which is to improve the accuracy of the intelligent processing of drug comparison data in a cardiac rehabilitation process.

[0005] To achieve the above-mentioned purpose, the present invention provides an intelligent processing system for drug comparison data in a cardiac rehabilitation process, comprising:

[0006] A data mining module is used to obtain the traditional Chinese medicine data in the cardiac rehabilitation process, wherein the traditional Chinese medicine data includes medicine formula data and medicine efficacy data, extract the core medicine data corresponding to the medicine formula data, and mine the medicine active ingredients and pharmacological information in the core medicine data;

[0007] A pharmacological analysis module, used to classify the pharmacological information to obtain the classified pharmacological information, extract the pharmacological mechanism information corresponding to the classified pharmacological information, and analyze the pharmacological similarity between the active ingredients of the medicaments in combination with the pharmacological mechanism information;

[0008] A drug efficacy analysis module, used to extract drug efficacy labels corresponding to the drug efficacy data, construct a label matrix corresponding to the drug efficacy data according to the drug efficacy labels, and analyze the drug efficacy correlation between the drug efficacy data according to the label matrix;

[0009] a drug interaction analysis module, for identifying the treatment cycle corresponding to the TCM drug data in the cardiac rehabilitation process, screening out the target drugs of the cycle from the TCM drug data according to the treatment cycle, and analyzing the drug interactions between the target drugs of the cycle in combination with the pharmacological similarity and the drug efficacy correlation;

[0010] The result generating module is used for combining the drug interaction and the periodic target drug to generate drug comparison results in the cardiac rehabilitation process.

[0011] Optionally, extracting the core pharmaceutical data corresponding to the pharmaceutical formula data includes:

[0012] Performing data cleaning processing on the pharmaceutical formula data to obtain cleaned pharmaceutical data;

[0013] Identify the medicine name in the cleaning medicine data, and count the medicine frequency of the medicine name in the cleaning medicine data;

[0014] Calculating the similarity between the drug name and the preset core drug table;

[0015] The core medicine data in the cleaning medicine data is extracted by combining the medicine frequency and the product name similarity.

[0016] Optionally, the calculating the similarity between the drug name and a preset core drug table includes:

[0017] Respectively extract the drug name and the product name characters in the preset core drug table to obtain the first product name character and the second product name character;

[0018] Standardizing the first product name characters and the second product name characters respectively to obtain first standard characters and second standard characters;

[0019] Performing vectorization processing on the first standard character and the second standard character respectively to obtain a first character vector and a second character vector;

[0020] Calculating vector similarity between the first character vector and the second character vector;

[0021] According to the vector similarity, the similarity between the drug name and the preset core drug table is obtained.

[0022] Optionally, the classifying the pharmacological information to obtain the classified pharmacological information includes:

[0023] Calculating an information entropy value corresponding to the pharmacological information, and extracting key pharmacological information from the pharmacological information according to the information entropy value;

[0024] Calculating the information edge probability and information union probability corresponding to the key pharmacological information;

[0025] Calculate the information correlation between the key pharmacological information by combining the information edge probability and the information association probability;

[0026] When the information correlation is greater than a preset correlation, the pharmacological information is classified to obtain classified pharmacological information.

[0027] Optionally, the combining the information edge probability and the information association probability to calculate the information association between the key pharmacological information includes:

[0028] Measure the amount of information corresponding to the key pharmacological information;

[0029] Combining the information edge probability, the information association probability and the information quantity, the information correlation between the key pharmacological information is calculated by the following formula:

[0030] ;

[0031] Among them, A represents the information correlation between key pharmacological information, represents the probability of information association corresponding to the cth information and the dth information in the key pharmacological information, and They respectively represent the information edge probabilities corresponding to the c-th information and the d-th information in the key pharmacological information, and c and d represent the serial numbers corresponding to the key pharmacological information.

[0032] Optionally, the combining of the pharmacological mechanism information to analyze the pharmacological similarity between the active ingredients of the medicaments includes:

[0033] Performing semantic analysis on the pharmacological mechanism information to obtain mechanism semantics, and extracting key semantic characters in the mechanism semantics;

[0034] According to the key semantic characters, construct similarity measurement indicators corresponding to the active ingredients of the medicine, and assign indicator weights corresponding to the similarity measurement indicators;

[0035] In combination with the pharmacological mechanism information, calculating the indicator similarity corresponding to the similarity measurement indicator;

[0036] Calculate the indicator similarity score between the active ingredients of the pharmaceutical agent by combining the indicator similarity and the indicator weight;

[0037] The pharmacological similarity between the active ingredients of the medicaments is analyzed based on the similarity scores of the indicators.

[0038] Optionally, constructing a label matrix corresponding to the drug efficacy data according to the drug efficacy label includes:

[0039] Deduplication processing is performed on the pharmacodynamic labels to obtain target pharmacodynamic labels;

[0040] Calculating the label interaction weights between the target pharmacodynamic labels;

[0041] Extracting characteristic labels from the target drug efficacy labels according to the label interaction weights;

[0042] Calculate the label feature value corresponding to the feature label;

[0043] According to the label feature values, a label matrix corresponding to the drug efficacy data is constructed.

[0044] Optionally, analyzing the pharmacodynamic correlation between the pharmacological efficacy data according to the label matrix includes:

[0045] Performing bidirectional label encoding processing on the label matrix to obtain a forward label encoding and a backward label encoding;

[0046] Performing encoding fusion processing on the forward label code and the backward label code to obtain a fused label code;

[0047] Calculating the correlation coefficient between the fusion label codes;

[0048] The drug efficacy correlation between the drug efficacy data is analyzed based on the correlation coefficient.

[0049] Optionally, performing encoding fusion processing on the forward label encoding and the backward label encoding to obtain a fused label encoding includes:

[0050] The forward label coding and the backward label coding are fused by the following formula:

[0051] ;

[0052] in, represents the fusion label encoding, represents the activation function, represents the xth code in the forward label encoding, represents the yth code in the backward label code, and x and y represent the serial numbers corresponding to the forward label code and the backward label code, respectively.

[0053] An intelligent processing method for drug comparison data in a cardiac rehabilitation process, characterized in that the method comprises:

[0054] Acquire traditional Chinese medicine data in the cardiac rehabilitation process, the traditional Chinese medicine data including: medicine formula data and medicine efficacy data, extract core medicine data corresponding to the medicine formula data, and mine out medicine active ingredients and pharmacological information in the core medicine data;

[0055] Classifying the pharmacological information to obtain classified pharmacological information, extracting pharmacological mechanism information corresponding to the classified pharmacological information, and analyzing the pharmacological similarity between the active ingredients of the medicaments in combination with the pharmacological mechanism information;

[0056] Extracting drug efficacy labels corresponding to the drug efficacy data, constructing a label matrix corresponding to the drug efficacy data according to the drug efficacy labels, and analyzing drug efficacy correlations between the drug efficacy data according to the label matrix;

[0057] Identify the treatment cycle corresponding to the TCM medicine data in the cardiac rehabilitation process, screen out the cycle target medicines from the TCM medicine data according to the treatment cycle, and analyze the drug interactions between the cycle target medicines in combination with the pharmacological similarity and the drug efficacy correlation;

[0058] The drug interaction and the periodic target drug are combined to generate drug comparison results during cardiac rehabilitation.

[0059] The present invention extracts the core drug data corresponding to the drug formula data, and can obtain key drug-related data in the drug formula data, and then can remove redundant data, so as to facilitate the subsequent mining of drug active ingredients and pharmacological information in the core drug data. The present invention classifies the pharmacological information, and can organize and sort the pharmacological information according to a certain classification system, so that the information is more systematic and structured. The pharmacological mechanism information corresponding to the classified pharmacological information is extracted, and the descriptive information of the specific pathways and methods of the drug corresponding to each category in the classified pharmacological information to produce therapeutic effects in the organism can be obtained. The present invention constructs a label matrix corresponding to the drug efficacy data based on the pharmacological label, so that the data is easier to understand and analyze. The visualization of the label matrix can intuitively observe the differences and trends in the efficacy of different drugs, so as to facilitate the subsequent analysis of the drug efficacy correlation corresponding to the drug efficacy data. The present invention selects the cycle target drug from the traditional Chinese medicine data according to the treatment cycle, and can obtain the treatment drug corresponding to each treatment cycle in the cardiac rehabilitation process, thereby facilitating the subsequent analysis of the drug interaction between the cycle target drugs. The present invention combines the drug interaction and the cycle target drug to generate the drug comparison results in the cardiac rehabilitation process, thereby improving the processing efficiency of the traditional Chinese medicine data in the cardiac rehabilitation process, and further provides a basis for the subsequent drug repositioning, drug design and drug combination research of cardiac rehabilitation treatment, so as to improve the efficiency of cardiac rehabilitation treatment. Therefore, the intelligent processing system and method for drug comparison data in the cardiac rehabilitation process provided by the embodiment of the present invention can improve the accuracy of intelligent processing of drug comparison data in the cardiac rehabilitation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 A functional module diagram of an intelligent processing system for drug comparison data in a cardiac rehabilitation process provided by an embodiment of the present invention;

[0061] Figure 2 A flowchart of an intelligent processing method for drug comparison data in a cardiac rehabilitation process provided by one embodiment of the present invention.

[0062] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0064] In addition, the step sequence in the following method embodiments is only an example and not a strict limitation.

[0065] In fact, the server-side device deployed by the intelligent processing system for comparing drug data in the cardiac rehabilitation process may be composed of one or more devices. The above-mentioned intelligent processing system for comparing drug data in the cardiac rehabilitation process can be implemented as: a business instance, a virtual machine, and a hardware device. For example, the intelligent processing system for comparing drug data in the cardiac rehabilitation process can be implemented as a business instance deployed on one or more devices in the cloud node. In simple terms, the live broadcast service system can be understood as a software deployed on the cloud node, which is used to provide intelligent processing services for comparing drug data in the cardiac rehabilitation process to each user terminal. Alternatively, the intelligent processing system for comparing drug data in the cardiac rehabilitation process can also be implemented as a virtual machine deployed on one or more devices in the cloud node. Application software for managing each user terminal is installed in the virtual machine. Alternatively, the intelligent processing system for comparing drug data in the cardiac rehabilitation process can also be implemented as a server composed of many hardware devices of the same or different types, and one or more hardware devices are set to provide intelligent processing services for comparing drug data in the cardiac rehabilitation process to each user terminal.

[0066] In terms of implementation, the intelligent processing system for comparing drug data during cardiac rehabilitation and the user end are adapted to each other. That is, the intelligent processing system for comparing drug data during cardiac rehabilitation is an application installed on a cloud service platform, and the user end is a client that establishes a communication connection with the application; or the intelligent processing system for comparing drug data during cardiac rehabilitation is implemented as a website, and the user end is implemented as a web page; or the intelligent processing system for comparing drug data during cardiac rehabilitation is implemented as a cloud service platform, and the user end is implemented as a small program in an instant messaging application.

[0067] Reference Figure 1 , which is a functional module diagram of an intelligent processing system for drug comparison data in a cardiac rehabilitation process provided by an embodiment of the present invention.

[0068] The intelligent processing system 100 for comparing drug data in the cardiac rehabilitation process of the present invention can be set in a cloud server. In terms of implementation, it can be used as one or more service devices, or as an application installed on the cloud (such as a server or server cluster of a live broadcast service operator), or it can also be developed as a website. According to the functions implemented, the intelligent processing system 100 for comparing drug data in the cardiac rehabilitation process includes a data mining module 101, a pharmacological analysis module 102, a drug efficacy analysis module 103, a drug interaction analysis module 104, and a result generation module 105.

[0069] In the embodiment of the present invention, in the tracking of the intelligent processing of the drug comparison data during the cardiac rehabilitation process, each of the above modules can be independently implemented and called with other modules. The call here can be understood as that a certain module can connect to multiple modules of another type and provide corresponding services to the multiple modules connected to it. In the intelligent processing system for the drug comparison data during the cardiac rehabilitation process provided by the embodiment of the present invention, the scope of application of the intelligent processing architecture for the drug comparison data during the cardiac rehabilitation process can be adjusted by adding modules and directly calling them without modifying the program code, so as to achieve cluster-type horizontal expansion, so as to achieve the purpose of quickly and flexibly expanding the intelligent processing system for the drug comparison data during the cardiac rehabilitation process. In actual applications, the above modules can be set in the same device or different devices, or they can be set in virtual devices, such as service instances in cloud servers.

[0070] In the following, in conjunction with specific embodiments, various components and specific workflows of the intelligent processing system for drug comparison data during cardiac rehabilitation are described.

[0071] The data mining module 101 is used to obtain the traditional Chinese medicine data in the cardiac rehabilitation process, the traditional Chinese medicine data including: medicine formula data and medicine efficacy data, extract the core medicine data corresponding to the medicine formula data, and mine the medicine active ingredients and pharmacological information in the core medicine data.

[0072] The present invention extracts the core drug data corresponding to the drug formula data, and can obtain key drug-related data in the drug formula data, and then can remove redundant data, so as to facilitate the subsequent mining of drug active ingredients and pharmacological information in the core drug data, wherein the traditional Chinese medicine data is the traditional Chinese medicine medication data used in the process of cardiac rehabilitation, the drug formula data is the specific medication data in the traditional Chinese medicine data, such as medication dosage and drug type, the drug efficacy data is the record data of the drug efficacy in the traditional Chinese medicine data, and the core drug data is the most used drug record data in the drug formula data.

[0073] As an embodiment of the present invention, the extracting of core medicine data corresponding to the medicine formula data includes: performing data cleaning processing on the medicine formula data to obtain cleaning medicine data, identifying medicine names in the cleaning medicine data, counting the medicine frequency of the medicine name in the cleaning medicine data, calculating the similarity between the medicine name and the preset core medicine table, and extracting the core medicine data in the cleaning medicine data by combining the medicine frequency and the product name similarity.

[0074] Among them, the cleaned drug data is the data obtained after cleaning the data corresponding to the invalid and missing values ​​in the drug formula data, the drug product name is the drug name corresponding to the cleaned drug data, the drug frequency is the number of times the drug product name appears in the cleaned drug data, the preset core drug table is a table or list defined in advance when marking drug formula data, which lists various possible core drug data types, and the product name similarity indicates the degree of similarity between the drug product name and the preset core drug table.

[0075] Optionally, the data cleaning processing of the pharmaceutical formula data can be implemented by a cleaning tool, and the cleaning tool is compiled by a scripting language. Identifying the pharmaceutical product name in the cleaned pharmaceutical data can be implemented by an OCR recognition method, and counting the pharmaceutical product frequency of the pharmaceutical product name in the cleaned pharmaceutical data can be implemented by a simple frequency calculation method. When the similarity between the pharmaceutical frequency and the product name is respectively greater than the corresponding preset thresholds, an extraction function is used to extract the core pharmaceutical data in the cleaned pharmaceutical data. The preset threshold corresponding to the similarity between the pharmaceutical frequency and the product name can be 0.8, and can also be set according to the actual business scenario. The extraction function includes a re.search() function.

[0076] Optionally, as an optional embodiment of the present invention, the calculating the similarity between the drug name and the preset core drug table includes: respectively extracting the drug name and the product name characters in the preset core drug table to obtain a first product name character and a second product name character, respectively standardizing the first product name character and the second product name character to obtain a first standard character and a second standard character, respectively vectorizing the first standard character and the second standard character to obtain a first character vector and a second character vector, calculating the vector similarity between the first character vector and the second character vector, and obtaining the product name similarity between the drug name and the preset core drug table based on the vector similarity.

[0077] Among them, the first product name character and the second product name character are respectively the words corresponding to the drug name and the drug name corresponding to the preset core drug table, the first standard character and the second standard character are respectively the characters obtained by removing special symbols and letters in the first product name character and the second product name character, the first character vector and the second character vector are respectively the expression vectors corresponding to the first standard character and the second standard character, and the vector similarity represents the degree of similarity between the first character vector and the second character vector.

[0078] Optionally, the above-mentioned OCR recognition technology can be used to extract the product name of the drug and the product name characters in the preset core drug table. The standardization processing of the first product name characters and the second product name characters can be achieved through string operation functions, such as strip() function and replace() function. The strip() function is used to remove spaces or specified characters at both ends of the string, and the replace() function is to replace specified characters or letters in the string. The vectorization processing of the first standard characters and the second standard characters can be achieved through the word2vec algorithm, and the calculation of the vector similarity between the first character vector and the second character vector can be achieved through the cosine similarity algorithm.

[0079] The present invention mines out the active ingredients and pharmacological information of the drugs in the core drug data to facilitate the subsequent analysis and processing of pharmacological similarity, wherein the active ingredients of the drugs are the main chemical components of the drugs in the core drug data, and the pharmacological information is the information on the pharmacological properties, mechanism of action and impact on biological systems of the drugs in the core drug data, involving various processes of interaction between drugs and organisms, including drug absorption, distribution, metabolism and excretion, as well as drug pharmacodynamics, side effects, interactions and other aspects. Optionally, the active ingredients of the drugs can be mined from the core drug data based on the data corresponding to the dosage and proportion of the drugs in the core drug data, and the pharmacological information corresponding to the active ingredients of the drugs, such as mechanism of action, pharmacodynamic properties, side effects, interactions and other information, can be queried from existing pharmacology databases, literature research and other resources through human-computer interaction.

[0080] The pharmacological analysis module 102 is used to classify the pharmacological information to obtain classified pharmacological information, extract the pharmacological mechanism corresponding to the classified pharmacological information, and analyze the pharmacological similarity between the active ingredients of the medicaments by combining the pharmacological mechanism and the classified pharmacological information.

[0081] The present invention classifies the pharmacological information, so that the pharmacological information can be organized and sorted according to a certain classification system, making the information more systematic and structured, and the pharmacological mechanism information corresponding to the classified pharmacological information can be extracted to obtain the description information of the specific ways and methods for the drugs corresponding to each category in the classified pharmacological information to produce therapeutic effects in vivo, wherein the classified pharmacological information is the information obtained after the pharmacological information is sorted according to a certain classification system, and the pharmacological mechanism information is the description information of the specific ways and methods for the drugs corresponding to each category in the classified pharmacological information to produce therapeutic effects in vivo. Optionally, the extraction of the pharmacological mechanism information corresponding to the classified pharmacological information can be achieved through an information search engine, and the information search engine is compiled by a scripting language.

[0082] As an embodiment of the present invention, the pharmacological information is classified to obtain classified pharmacological information, including: calculating the information entropy value corresponding to the pharmacological information, extracting key pharmacological information from the pharmacological information according to the information entropy value, calculating the information edge probability and the information association probability corresponding to the key pharmacological information, calculating the information correlation between the key pharmacological information by combining the information edge probability and the information association probability, and when the information correlation is greater than a preset correlation, classifying the pharmacological information to obtain classified pharmacological information.

[0083] Among them, the information entropy value represents a measure of the uncertainty corresponding to the pharmacological information, the lower the value means the more certain the result of the random variable, and the higher the value means the greater the uncertainty of the result. The key pharmacological information is the important information in the pharmacological information. The information edge probability and the information joint probability respectively represent the frequency of occurrence of each pharmacological information in the key pharmacological information and the joint frequency of the simultaneous occurrence of two pharmacological information. The information correlation degree represents the degree of correlation between the key pharmacological information. The preset correlation degree is the judgment standard corresponding to the information correlation degree, which can be set according to the specific application scenario.

[0084] Optionally, the information entropy value corresponding to the pharmacological information can be calculated by using the Shannon entropy formula, the minimum entropy value among the information entropy values ​​is identified, and the pharmacological information corresponding to the minimum entropy value is extracted to obtain key pharmacological information. The information edge probability and the information union probability can be obtained by counting the number of occurrences of each information in the key pharmacological information and the number of simultaneous occurrences of two information, dividing the two numbers by the total number of information and multiplying by 100%, and classifying the pharmacological information.

[0085] Optionally, as an optional embodiment of the present invention, the combining the information edge probability and the information association probability to calculate the information association between the key pharmacological information includes: measuring the number of information corresponding to the key pharmacological information, combining the information edge probability, the information association probability and the number of information, and calculating the information association between the key pharmacological information by the following formula:

[0086] ;

[0087] Among them, A represents the information correlation between key pharmacological information, represents the probability of information association corresponding to the cth information and the dth information in the key pharmacological information, and They respectively represent the information edge probabilities corresponding to the c-th information and the d-th information in the key pharmacological information, and c and d represent the serial numbers corresponding to the key pharmacological information.

[0088] The present invention analyzes the pharmacological similarity between the active ingredients of the medicaments by combining the pharmacological mechanism and the classified pharmacological information, thereby understanding the degree of similarity in efficacy between the active ingredients of the medicaments and providing useful guidance and suggestions for alternative treatments, drug combination treatments and side effect prediction, wherein the pharmacological similarity represents the degree of similarity between the active ingredients of the medicaments.

[0089] As an embodiment of the present invention, the method of analyzing the pharmacological similarity between the active ingredients of the medicines in combination with the pharmacological mechanism information includes: performing semantic analysis on the pharmacological mechanism information to obtain mechanism semantics, extracting key semantic characters in the mechanism semantics, constructing similarity measurement indicators corresponding to the active ingredients of the medicines based on the key semantic characters, assigning indicator weights corresponding to the similarity measurement indicators, calculating the indicator similarity corresponding to the similarity measurement indicators in combination with the pharmacological mechanism information, calculating the indicator similarity score between the active ingredients of the medicines in combination with the indicator similarity and the indicator weight, and analyzing the pharmacological similarity between the active ingredients of the medicines based on the indicator similarity score.

[0090] Among them, the mechanism semantics is the explanation of the meaning corresponding to the pharmacological mechanism information, the key semantic characters are representative characters in the mechanism semantics, the similarity measurement index is the index for similarity analysis of the active ingredients of the pharmaceutical agent, the index weight represents the index importance corresponding to the similarity measurement index, the index similarity represents the sum of the information similarities corresponding to each similarity measurement index in the pharmacological mechanism information, and the index similarity score represents the similarity score value between the active ingredients of the pharmaceutical agent.

[0091] Optionally, semantic analysis of the pharmacological mechanism information can be achieved through a semantic analysis method, and extraction of key semantic characters in the mechanism semantics can be achieved through the above-mentioned OCR recognition technology. The key semantic characters are used as similarity measurement indicators corresponding to the effective ingredients of the drug, such as targets, pharmacological activity, etc., by collecting historical treatment data during cardiac rehabilitation, and using correlation analysis, principal component analysis and other methods based on the historical treatment data to evaluate the degree of influence of different indicators on drug characteristics and application scenarios, and quantify the degree of influence into a numerical value, which is used as the indicator weight corresponding to the similarity measurement indicator, and the The pharmacological mechanism information corresponding to the similarity measurement index is normalized to obtain the index normalized information, the information similarity between the index normalized information is calculated, the information similarity is summed to obtain the index similarity, the index similarity and the index weight are multiplied and summed to obtain the index similarity score between the active ingredients of the pharmaceutical agents, the index similarity score is compared with a preset score interval, and the pharmacological similarity between the active ingredients of the pharmaceutical agents is analyzed. For example, if the score interval is high similarity: 8-10 points, medium similarity: 6-8 points, and the index similarity score is 9 points, it means that the pharmacological similarity between the active ingredients of the pharmaceutical agents is high.

[0092] The pharmacodynamic analysis module 103 is used to extract the pharmacodynamic labels corresponding to the drug efficacy data, construct a label matrix corresponding to the drug efficacy data according to the pharmacodynamic labels, and analyze the pharmacodynamic correlation between the drug efficacy data according to the label matrix.

[0093] The present invention constructs a label matrix corresponding to the drug efficacy data based on the drug efficacy label, so that the data is easier to understand and analyze. By visualizing the label matrix, the differences and trends in efficacy of different drugs can be intuitively observed, so as to facilitate the subsequent analysis of the drug efficacy correlation corresponding to the drug efficacy data. The drug efficacy label is a label in the drug efficacy data that describes and evaluates the drug efficacy, and the label matrix is ​​a square matrix constructed according to the drug efficacy label for each data in the drug efficacy data. Optionally, extracting the drug efficacy label corresponding to the drug efficacy data can be achieved by a label extraction tool, and the label extraction tool is compiled by a scripting language.

[0094] As an embodiment of the present invention, constructing a label matrix corresponding to the drug efficacy data based on the drug efficacy label includes: deduplicating the drug efficacy label to obtain a target drug efficacy label, calculating the label interaction weights between the target drug efficacy labels, extracting feature labels in the target drug efficacy labels based on the label interaction weights, calculating the label feature values ​​corresponding to the feature labels, and constructing a label matrix corresponding to the drug efficacy data based on the label feature values.

[0095] Among them, the target pharmacodynamic label is the label obtained after removing duplicate labels in the pharmacodynamic label, the label interaction weight represents the mutual dependence between the target pharmacodynamic labels, the feature label is the representative label in the target pharmacodynamic label, and the label feature value is the specific numerical value corresponding to the feature label.

[0096] Optionally, deduplication of the pharmacodynamic labels can be achieved through the drop_duplicates() function, and the Pearson correlation coefficient between the target pharmacodynamic labels can be calculated to obtain label interaction weights, where the Pearson correlation coefficient represents the strength of the linear relationship between the target pharmacodynamic labels. The label with the highest weight value in the label interaction weights is extracted as a feature label, and the feature label vector can be obtained by performing vector processing on the feature label, and the vector eigenvalue corresponding to the feature label vector is calculated to obtain the label eigenvalue, and the label matrix corresponding to the drug efficacy data can be constructed through a matrix function, such as a zero matrix function.

[0097] The present invention analyzes the pharmacodynamic correlation between the drug efficacy data based on the label matrix, and then can understand the correlation between the drug efficacy data, thereby facilitating the subsequent analysis of the drug interaction between the periodic target drugs, wherein the pharmacodynamic correlation represents the pharmacodynamic correlation between the drug efficacy data.

[0098] As an embodiment of the present invention, the method of analyzing the pharmacodynamic correlation between the drug efficacy data based on the label matrix includes: performing bidirectional label encoding processing on the label matrix to obtain forward label encoding and backward label encoding, performing encoding fusion processing on the forward label encoding and the backward label encoding to obtain fused label encoding, calculating the correlation coefficient between the fused label encodings, and analyzing the pharmacodynamic correlation between the drug efficacy data based on the correlation coefficient.

[0099] Among them, the forward label coding and the backward label coding are the label codings obtained after forward and backward encoding of the label matrix respectively, the fused label coding is the label coding obtained after the forward label coding and the backward label coding are fused, and the correlation coefficient represents the degree of association between the fused label codings.

[0100] Optionally, bidirectional label encoding processing of the label matrix can be implemented through a bidirectional recurrent neural network. The standard deviation corresponding to each code in the fused label code can be calculated, and the coding covariance between the codes in the fused label code can be calculated. The coding covariance is divided by the standard deviation corresponding to the corresponding code to obtain the correlation coefficient between the fused label codes. If the correlation coefficient is a positive number, it indicates that the pharmacodynamic correlation between the drug efficacy data is positively correlated. If the correlation coefficient is a negative number, it indicates that the pharmacodynamic correlation between the drug efficacy data is positively correlated.

[0101] Optionally, as an optional embodiment of the present invention, the performing encoding fusion processing on the forward label encoding and the backward label encoding to obtain the fused label encoding includes:

[0102] The forward label coding and the backward label coding are fused by the following formula:

[0103] ;

[0104] in, represents the fusion label encoding, represents the activation function, represents the xth code in the forward label encoding, represents the yth code in the backward label encoding, x and y represent the serial numbers corresponding to the forward label encoding and the backward label encoding respectively.

[0105] The drug interaction analysis module 104 is used to identify the treatment cycle corresponding to the traditional Chinese medicine data in the cardiac rehabilitation process, screen out the cycle target drugs from the traditional Chinese medicine data according to the treatment cycle, and analyze the drug interactions between the cycle target drugs in combination with the pharmacological similarity and the drug efficacy correlation.

[0106] The present invention screens out cycle target drugs from the traditional Chinese medicine data according to the treatment cycle, and can obtain the treatment drugs corresponding to each treatment cycle in the cardiac rehabilitation process, thereby facilitating the subsequent analysis of drug interactions between the cycle target drugs, wherein the treatment cycle is the corresponding treatment time period in the cardiac rehabilitation process, such as the early or middle stage, and the cycle target drug is the drug used in the cardiac rehabilitation process corresponding to the traditional Chinese medicine data within the treatment cycle. Optionally, the time recording data in the traditional Chinese medicine data in the cardiac rehabilitation process is identified, and the treatment cycle corresponding to the traditional Chinese medicine data is determined through the time recording data.

[0107] The present invention analyzes the drug interaction between the cycle target drugs by combining the pharmacological similarity and the pharmacodynamic correlation, and then understands the mutual influence between the cycle target drugs, so as to facilitate the formulation of subsequent drug formulations, wherein the drug interaction is the mutual influence between the cycle target drugs. Optionally, the drug interaction between the cycle target drugs can be analyzed by combining the high and low pharmacological similarity and the pharmacodynamic correlation. For example, if the pharmacological similarity and the pharmacodynamic correlation are both high, the drug interaction between the cycle target drugs is synergistic or enhancing. If the pharmacological similarity and the pharmacodynamic correlation are both low, the drug interaction between the cycle target drugs is inhibitory or non-synergistic.

[0108] The result generation module 105 is used to generate a drug comparison result in the cardiac rehabilitation process by combining the drug interaction and the periodic target drug.

[0109] The present invention generates drug comparison results in the cardiac rehabilitation process by combining the drug interaction and the periodic target drug, thereby improving the processing efficiency of the traditional Chinese medicine data in the cardiac rehabilitation process, and further providing a basis for subsequent drug repositioning, drug design and drug combination research in cardiac rehabilitation treatment, so as to improve the efficiency of cardiac rehabilitation treatment.

[0110] The present invention extracts the core drug data corresponding to the drug formula data, and can obtain key drug-related data in the drug formula data, and then can remove redundant data, so as to facilitate the subsequent mining of drug active ingredients and pharmacological information in the core drug data. The present invention classifies the pharmacological information, and can organize and sort the pharmacological information according to a certain classification system, so that the information is more systematic and structured. The pharmacological mechanism information corresponding to the classified pharmacological information is extracted, and the descriptive information of the specific pathways and methods of the drug corresponding to each category in the classified pharmacological information to produce therapeutic effects in the organism can be obtained. The present invention constructs a label matrix corresponding to the drug efficacy data based on the pharmacological label, so that the data is easier to understand and analyze. The visualization of the label matrix can intuitively observe the differences and trends in the efficacy of different drugs, so as to facilitate the subsequent analysis of the drug efficacy correlation corresponding to the drug efficacy data. The present invention screens out the cycle target drugs from the traditional Chinese medicine data according to the treatment cycle, and can obtain the treatment drugs corresponding to each treatment cycle in the cardiac rehabilitation process, thereby facilitating the subsequent analysis of the drug interaction between the cycle target drugs. The present invention generates drug comparison results in the cardiac rehabilitation process by combining the drug interaction and the cycle target drugs, thereby improving the processing efficiency of the traditional Chinese medicine data in the cardiac rehabilitation process, and further providing a basis for the subsequent drug repositioning, drug design and drug combination research of cardiac rehabilitation treatment, so as to improve the efficiency of cardiac rehabilitation treatment. Therefore, the intelligent processing system for drug comparison data in the cardiac rehabilitation process provided by the embodiment of the present invention can improve the accuracy of intelligent processing of drug comparison data in the cardiac rehabilitation process.

[0111] Reference Figure 2 FIG. 1 is a flow chart of a method for intelligently processing drug comparison data in a cardiac rehabilitation process according to an embodiment of the present invention. In this embodiment, the method for intelligently processing drug comparison data in a cardiac rehabilitation process includes:

[0112] Acquire traditional Chinese medicine data in the cardiac rehabilitation process, the traditional Chinese medicine data including: medicine formula data and medicine efficacy data, extract core medicine data corresponding to the medicine formula data, and mine out medicine active ingredients and pharmacological information in the core medicine data;

[0113] Classifying the pharmacological information to obtain classified pharmacological information, extracting pharmacological mechanism information corresponding to the classified pharmacological information, and analyzing the pharmacological similarity between the active ingredients of the medicaments in combination with the pharmacological mechanism information;

[0114] Extracting drug efficacy labels corresponding to the drug efficacy data, constructing a label matrix corresponding to the drug efficacy data according to the drug efficacy labels, and analyzing drug efficacy correlations between the drug efficacy data according to the label matrix;

[0115] Identify the treatment cycle corresponding to the TCM medicine data in the cardiac rehabilitation process, screen out the cycle target medicines from the TCM medicine data according to the treatment cycle, and analyze the drug interactions between the cycle target medicines in combination with the pharmacological similarity and the drug efficacy correlation;

[0116] The drug interaction and the periodic target drug are combined to generate drug comparison results during cardiac rehabilitation.

[0117] In the several embodiments provided by the present invention, it should be understood that the provided system and method can be implemented in other ways. For example, the system embodiment described above is only illustrative, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0118] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0119] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. An intelligent processing system for drug comparison data in cardiac rehabilitation process, characterized in that: The system comprises: A data mining module is used to obtain the traditional Chinese medicine data in the cardiac rehabilitation process, wherein the traditional Chinese medicine data includes medicine formula data and medicine efficacy data, extract the core medicine data corresponding to the medicine formula data, and mine the medicine active ingredients and pharmacological information in the core medicine data; A pharmacological analysis module, used to classify the pharmacological information to obtain the classified pharmacological information, extract the pharmacological mechanism information corresponding to the classified pharmacological information, and analyze the pharmacological similarity between the active ingredients of the medicaments in combination with the pharmacological mechanism information; A drug efficacy analysis module, used to extract drug efficacy labels corresponding to the drug efficacy data, construct a label matrix corresponding to the drug efficacy data according to the drug efficacy labels, and analyze the drug efficacy correlation between the drug efficacy data according to the label matrix; a drug interaction analysis module, for identifying the treatment cycle corresponding to the TCM drug data in the cardiac rehabilitation process, screening out the target drugs of the cycle from the TCM drug data according to the treatment cycle, and analyzing the drug interactions between the target drugs of the cycle in combination with the pharmacological similarity and the drug efficacy correlation; The result generating module is used for combining the drug interaction and the periodic target drug to generate drug comparison results in the cardiac rehabilitation process.

2. The intelligent processing system for drug comparison data in cardiac rehabilitation process according to claim 1, characterized in that: The extracting the core pharmaceutical data corresponding to the pharmaceutical formula data includes: Performing data cleaning processing on the pharmaceutical formula data to obtain cleaned pharmaceutical data; Identify the medicine name in the cleaning medicine data, and count the medicine frequency of the medicine name in the cleaning medicine data; Calculating the similarity between the drug name and the preset core drug table; The core medicine data in the cleaning medicine data is extracted by combining the medicine frequency and the product name similarity.

3. The intelligent processing system for drug comparison data in cardiac rehabilitation process according to claim 2, characterized in that: The calculating the similarity between the drug name and the preset core drug table includes: Respectively extract the drug name and the product name characters in the preset core drug table to obtain the first product name character and the second product name character; Standardizing the first product name characters and the second product name characters respectively to obtain first standard characters and second standard characters; Performing vectorization processing on the first standard character and the second standard character respectively to obtain a first character vector and a second character vector; Calculating vector similarity between the first character vector and the second character vector; According to the vector similarity, the similarity between the drug name and the preset core drug table is obtained.

4. The intelligent processing system for drug comparison data in cardiac rehabilitation process according to claim 1, characterized in that: The classifying process of the pharmacological information to obtain the classified pharmacological information includes: Calculating an information entropy value corresponding to the pharmacological information, and extracting key pharmacological information from the pharmacological information according to the information entropy value; Calculating the information edge probability and information union probability corresponding to the key pharmacological information; Calculate the information correlation between the key pharmacological information by combining the information edge probability and the information association probability; When the information correlation is greater than a preset correlation, the pharmacological information is classified to obtain classified pharmacological information.

5. The intelligent processing system for drug comparison data in cardiac rehabilitation process according to claim 4, characterized in that: The combining the information edge probability and the information association probability to calculate the information correlation between the key pharmacological information includes: Measure the amount of information corresponding to the key pharmacological information; Combining the information edge probability, the information association probability and the information quantity, the information correlation between the key pharmacological information is calculated by the following formula: ； Among them, A represents the information correlation between key pharmacological information, represents the probability of information association corresponding to the cth information and the dth information in the key pharmacological information, and They respectively represent the information edge probabilities corresponding to the c-th information and the d-th information in the key pharmacological information, and c and d represent the serial numbers corresponding to the key pharmacological information.

6. The intelligent processing system for drug comparison data in cardiac rehabilitation process according to claim 1, characterized in that: The combining of the pharmacological mechanism information to analyze the pharmacological similarity between the active ingredients of the medicaments includes: Performing semantic analysis on the pharmacological mechanism information to obtain mechanism semantics, and extracting key semantic characters in the mechanism semantics; According to the key semantic characters, construct similarity measurement indicators corresponding to the active ingredients of the medicine, and assign indicator weights corresponding to the similarity measurement indicators; In combination with the pharmacological mechanism information, calculating the indicator similarity corresponding to the similarity measurement indicator; Calculate the indicator similarity score between the active ingredients of the pharmaceutical agent by combining the indicator similarity and the indicator weight; The pharmacological similarity between the active ingredients of the medicaments is analyzed based on the similarity scores of the indicators.

7. The intelligent processing system for drug comparison data in cardiac rehabilitation process according to claim 1, characterized in that: The step of constructing a label matrix corresponding to the drug efficacy data according to the drug efficacy label includes: Deduplication processing is performed on the pharmacodynamic labels to obtain target pharmacodynamic labels; Calculating the label interaction weights between the target pharmacodynamic labels; Extracting characteristic labels from the target drug efficacy labels according to the label interaction weights; Calculate the label feature value corresponding to the feature label; According to the label feature values, a label matrix corresponding to the drug efficacy data is constructed.

8. The intelligent processing system for drug comparison data in cardiac rehabilitation process according to claim 1, characterized in that: Analyzing the pharmacodynamic correlation between the pharmacological efficacy data according to the label matrix includes: Performing bidirectional label encoding processing on the label matrix to obtain a forward label encoding and a backward label encoding; Performing encoding fusion processing on the forward label code and the backward label code to obtain a fused label code; Calculating the correlation coefficient between the fusion label codes; The drug efficacy correlation between the drug efficacy data is analyzed based on the correlation coefficient.

9. The intelligent processing system for drug comparison data in cardiac rehabilitation process according to claim 8, characterized in that: The performing encoding fusion processing on the forward label code and the backward label code to obtain the fused label code includes: The forward label coding and the backward label coding are fused by the following formula: ； in, represents the fusion label encoding, represents the activation function, represents the xth code in the forward label encoding, represents the yth code in the backward label code, and x and y represent the serial numbers corresponding to the forward label code and the backward label code, respectively.

10. An intelligent processing method for drug comparison data in cardiac rehabilitation process, characterized in that: The method comprises: Acquire traditional Chinese medicine data in the cardiac rehabilitation process, the traditional Chinese medicine data including: medicine formula data and medicine efficacy data, extract core medicine data corresponding to the medicine formula data, and mine out medicine active ingredients and pharmacological information in the core medicine data; Classifying the pharmacological information to obtain classified pharmacological information, extracting pharmacological mechanism information corresponding to the classified pharmacological information, and analyzing the pharmacological similarity between the active ingredients of the medicaments in combination with the pharmacological mechanism information; Extracting drug efficacy labels corresponding to the drug efficacy data, constructing a label matrix corresponding to the drug efficacy data according to the drug efficacy labels, and analyzing drug efficacy correlations between the drug efficacy data according to the label matrix; Identify the treatment cycle corresponding to the TCM medicine data in the cardiac rehabilitation process, screen out the cycle target medicines from the TCM medicine data according to the treatment cycle, and analyze the drug interactions between the cycle target medicines in combination with the pharmacological similarity and the drug efficacy correlation; The drug interaction and the periodic target drug are combined to generate drug comparison results during cardiac rehabilitation.

Images