Traditional Chinese medicine diagnosis auxiliary method and device and storage medium

Abstract

The invention discloses a traditional Chinese medicine diagnosis assisting method and device and a storage medium, and relates to the technical field of artificial intelligence, and the traditional Chinese medicine diagnosis assisting method comprises the steps: obtaining a target dependent variable selected by a target user; determining a plurality of first independent variable groups corresponding to the target dependent variable; performing iterative training on an initial dependent variable prediction model based on a preset training algorithm and each first independent variable group to obtain a plurality of to-be-determined dependent variable prediction models corresponding to the target dependent variable; and carrying out accuracy sorting on each undetermined dependent variable prediction model, and determining a target dependent variable prediction model so as to carry out diagnosis assistance on the target dependent variable. According to the method, a large amount of data is analyzed and summarized, and then iterative training is performed through large models of different algorithms, so that the problems of data dispersion, large data processing workload, difficulty in converting texts into achievements, difficulty in fusing AI into clinical decisions and the like in the medical industry are solved, and the convenience of seeing a doctor is improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular, to a method, device, and storage medium for assisting traditional Chinese medicine diagnosis. Background Art

[0002] At present, artificial intelligence has been applied to a certain extent in all walks of life and has reduced a large amount of workload. However, for the special medical industry, the application of artificial intelligence has been hindered. Especially for traditional Chinese medicine, there are the following problems in the actual application process. First, the data is scattered, the quality of the original data is not high, it is difficult to converge the data, and the text data and image data are separated, etc.; second, the data processing workload is large, the manual efficiency is low, and the data reuse rate is low; third, it is difficult to transform clinically, and it is difficult to transform the papers and patient data into clinical use, lacking continuous system iteration and system use feedback closed-loop; fourth, it is difficult for artificial intelligence technologies such as AI to be integrated into CDSS (Clinical Decision Support System), and it is impossible to provide a convenient medical experience for doctors and patients. Summary of the Invention

[0003] The main purpose of the present application is to provide a method, device, and storage medium for assisting traditional Chinese medicine diagnosis, aiming to solve problems such as scattered data, large data processing workload, difficulty in transforming text into results, and difficulty in integrating AI into clinical decision-making in the medical industry, and improve the convenience of medical treatment.

[0004] To achieve the above object, the present application proposes a method for assisting traditional Chinese medicine diagnosis, and the method includes:

[0005] Obtain a target dependent variable selected by a target user;

[0006] Determine a plurality of first independent variable groups corresponding to the target dependent variable;

[0007] Based on a preset training algorithm and each of the first independent variable groups, iteratively train an initial dependent variable prediction model to obtain a plurality of pending dependent variable prediction models corresponding to the target dependent variable;

[0008] Sort the accuracies of each of the pending dependent variable prediction models to determine a target dependent variable prediction model for assisting in the diagnosis of the target dependent variable.

[0009] In one embodiment, the determining a plurality of first independent variable groups corresponding to the target dependent variable includes:

[0010] Query the target dependent variable with a preset relational database to determine a plurality of second independent variables corresponding to the target dependent variable;

[0011] Classify each of the second independent variables to obtain a number of second independent variable groups;

[0012] Based on each of the second independent variable groups, determine a number of first independent variable groups corresponding to the target dependent variable.

[0013] In one embodiment, the determining, based on each of the second independent variable groups, a number of first independent variable groups corresponding to the target dependent variable includes:

[0014] Obtain the valid rate corresponding to each of the second independent variables;

[0015] Sort the valid rates of the second independent variables in each of the second independent variable groups, and eliminate the second independent variables that do not meet the preset valid rate range in each group;

[0016] Use the second independent variable groups after elimination as the first independent variable groups.

[0017] In one embodiment, the iteratively training the initial dependent variable prediction model based on a preset training algorithm and each of the first independent variable groups to obtain a number of pending dependent variable prediction models corresponding to the target dependent variable includes:

[0018] Associate and combine each of the first independent variable groups to obtain a number of groups of training samples;

[0019] Input each of the groups of training samples into the initial dependent variable prediction model applying different preset training algorithms for iterative training to obtain a number of pending dependent variable prediction models corresponding to the target dependent variable.

[0020] In one embodiment, the ranking the accuracies of each of the pending dependent variable prediction models to determine a target dependent variable prediction model for assisting in diagnosing the target dependent variable includes:

[0021] Make a horizontal comparison of each of the pending dependent variable prediction models to obtain a model comparison result;

[0022] Make a directional match of each of the pending dependent variable prediction models to obtain a model matching result;

[0023] Based on the model comparison result and the model matching result, determine the target dependent variable prediction model and the model ranking of the target dependent variable prediction model;

[0024] Push the target dependent variable prediction model to the target user according to the model ranking to assist in diagnosing the target dependent variable.

[0025] In one embodiment, the making a horizontal comparison of each of the pending dependent variable prediction models to obtain a model comparison result includes:

[0026] Obtain the stability evaluation results, accuracy, precision, and recall rate of each of the to-be-determined dependent variable prediction models;

[0027] Calculate the accuracy, precision, and recall rate to determine the accuracy evaluation result of the to-be-determined dependent variable prediction model;

[0028] Based on the stability evaluation result and the accuracy evaluation result, obtain the model comparison result.

[0029] In one embodiment, after ranking the accuracies of the to-be-determined dependent variable prediction models to determine the target dependent variable prediction model for diagnostic assistance for the target dependent variable, it further includes:

[0030] Obtain the user score of the target dependent variable prediction model;

[0031] If the user score meets the preset recommended score range, add the target dependent variable prediction model and the usage instructions corresponding to the target dependent variable prediction model to the recommended model library;

[0032] Push the target dependent variable prediction model to medical experts for model evaluation by the medical experts.

[0033] In one embodiment, before iteratively training the initial dependent variable prediction model based on the preset training algorithm and each of the first independent variable groups to obtain several to-be-determined dependent variable prediction models corresponding to the target dependent variable, it further includes:

[0034] Associate and combine the target dependent variable with each of the first independent variables to obtain variable combinations, and determine whether there are identical variable combinations;

[0035] If so, use the target dependent variable prediction model corresponding to the variable combination as the target dependent variable prediction model corresponding to the target dependent variable for diagnostic assistance for the target dependent variable.

[0036] In addition, to achieve the above object, the present application also proposes a traditional Chinese medicine medical diagnosis assistance device, and the traditional Chinese medicine medical diagnosis assistance device includes:

[0037] An acquisition module, configured to acquire a target dependent variable selected by a target user;

[0038] A determination module, configured to determine several first independent variable groups corresponding to the target dependent variable;

[0039] A training module, configured to iteratively train an initial dependent variable prediction model based on the preset training algorithm and each of the first independent variable groups to obtain several to-be-determined dependent variable prediction models corresponding to the target dependent variable;

[0040] A sorting module, configured to sort the accuracy of each of the to-be-determined dependent variable prediction models, determine a target dependent variable prediction model, and assist in diagnosis for the target dependent variable.

[0041] In addition, to achieve the above object, the present application further provides a traditional Chinese medicine diagnosis assistance device, which includes: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the computer program is configured to implement the steps of the traditional Chinese medicine diagnosis assistance method as described above.

[0042] In addition, to achieve the above object, the present application further provides a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium, and when the computer program is executed by a processor, the steps of the traditional Chinese medicine diagnosis assistance method as described above are implemented.

[0043] In addition, to achieve the above object, the present application further provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps of the traditional Chinese medicine diagnosis assistance method as described above are implemented.

[0044] The present application provides a traditional Chinese medicine diagnosis assistance method, device and storage medium. The traditional Chinese medicine diagnosis assistance method obtains a target dependent variable selected by a target user, then determines a plurality of first independent variable groups corresponding to the target dependent variable, and then iteratively trains an initial dependent variable prediction model based on a preset training algorithm and each of the first independent variable groups to obtain a plurality of to-be-determined dependent variable prediction models corresponding to the target dependent variable, and then sorts the accuracy of each of the to-be-determined dependent variable prediction models to determine a target dependent variable prediction model for assisting in diagnosis for the target dependent variable, thereby solving problems such as scattered data, large data processing workload, difficulty in converting text into results, and difficulty in integrating AI into clinical decision-making in the medical industry, and improving the convenience of medical treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The accompanying drawings here are incorporated into the specification and form a part of this specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0046] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0047] Figure 1It is a schematic flowchart provided for the first embodiment of the traditional Chinese medicine medical diagnosis assistance method in this application;

[0048] Figure 2 It is a schematic flowchart provided for the second embodiment of the traditional Chinese medicine medical diagnosis assistance method in this application;

[0049] Figure 3 It is a schematic flowchart provided for the third embodiment of the traditional Chinese medicine medical diagnosis assistance method in this application;

[0050] Figure 4 It is a schematic module structure diagram of the traditional Chinese medicine medical diagnosis assistance device in the embodiment of this application;

[0051] Figure 5 It is a schematic device structure diagram of the hardware operating environment involved in the traditional Chinese medicine medical diagnosis assistance method in the embodiment of this application.

[0052] The implementation, functional features and advantages of this application will be further described in combination with the embodiments with reference to the accompanying drawings. Specific Embodiments

[0053] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of this application and are not used to limit this application.

[0054] In order to better understand the technical solutions of this application, the following will be described in detail in combination with the accompanying drawings of the specification and specific embodiments.

[0055] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device, a big data service platform, a traditional Chinese medicine medical diagnosis assistance system, etc. that can implement the above functions. The following takes the traditional Chinese medicine medical diagnosis assistance system as an example to illustrate this embodiment and the following embodiments.

[0056] Based on this, the embodiment of this application provides a traditional Chinese medicine medical diagnosis assistance method, referring to Figure 1 , Figure 1 It is a schematic flowchart provided for the first embodiment of the traditional Chinese medicine medical diagnosis assistance method in this application.

[0057] In this embodiment, the traditional Chinese medicine medical diagnosis assistance method includes steps S11 to S14:

[0058] Step S11, obtaining the target dependent variable selected by the target user;

[0059] It should be noted that the target user refers to users who need to conduct medical research such as doctors, physicians and medical writers, etc., which are not limited here and can be defined according to actual situations.

[0060] Further, it should be noted that the target dependent variable refers to the metric that the target user wants to predict, and this metric can include one or more variables. For example, if an expert needs to establish a prediction model for the drug effect after using a certain prescription drug, then the target dependent variable is the metric that the target user wants to predict - the drug effect. Among them, the variables that the drug effect can include are whether sleep is improved, whether diet is improved, etc. There is no limitation here, and the dependent variable can be set according to the actual situation.

[0061] Specifically, it can be obtained by the target user selecting the target dependent variable from the preset metric database. Among them, the preset metric database refers to a pre-set database that stores multiple medical metrics related to healthcare. The medical metrics can be collected and stored from the diagnosis information of patients diagnosed by traditional Chinese medicine experts. For example, if the diagnosis information of a patient diagnosed by a traditional Chinese medicine expert is "dizziness and headache, tinnitus and brain ringing, feeling like being wrapped on the top of the head, fatigue, unsteady walking, hot flashes and night sweats, especially in the head and neck, dry mouth and bitter taste, restless sleep, irregular diet and bowel movements, first dry and then loose, three times a day, one time of nocturia, dark purple tongue, varicose veins under the tongue, enlarged tongue body with tooth marks on the sides, cracks in the middle, generally, thin yellow and greasy tongue coating, thready, slippery and rapid pulse", then through natural semantic recognition processing, different types of metrics are extracted from it, such as inspection: tongue inspection, facial complexion inspection, skin inspection; auscultation: sound listening, odor smelling, etc.; inquiry: diet and taste, sleep, defecation and urination, menstrual period of women; palpation: pulse condition and other medical metrics. Additionally, the preset metric database can also obtain metrics through data crawlers and other means. There is no limitation here, and it can be set according to the actual situation.

[0062] Step S12, determine a number of first independent variable groups corresponding to the target dependent variable;

[0063] It should be noted that the first independent variable group refers to the independent variable group obtained after classifying the independent variables corresponding to the target dependent variable. Continuing the above example, the inspection group can include independent variables such as tongue inspection, facial complexion inspection, and skin inspection.

[0064] Specifically, query the target dependent variable with the preset relationship database to determine a number of second independent variables corresponding to the target dependent variable, and then classify each of the second independent variables to obtain a number of second independent variable groups. Thus, based on each of the second independent variable groups, determine a number of first independent variable groups corresponding to the target dependent variable.

[0065] Step S13, based on the preset training algorithm and each of the first independent variable groups, perform iterative training on the initial dependent variable prediction model to obtain a number of pending dependent variable prediction models corresponding to the target dependent variable;

[0066] It should be noted that the preset training algorithms refer to various pre-stored model training algorithms, such as SVM (Support Vector Machine), LASSO (Least Absolute Shrinkage and Selection Operator), XGBoost (eXtreme Gradient Boosting), etc. There is no limitation here.

[0067] Furthermore, it should be noted that the initial dependent variable prediction model refers to an initial model that has not been trained, and the to-be-determined dependent variable prediction model refers to a dependent variable prediction model that has been trained but not evaluated.

[0068] Specifically, the first independent variable groups are associated and combined to obtain several groups of training samples. Then, each group of training samples is respectively input into the initial dependent variable prediction models applying different preset training algorithms for iterative training to obtain several to-be-determined dependent variable prediction models corresponding to the target dependent variable.

[0069] Step S14: Sort the to-be-determined dependent variable prediction models by accuracy to determine the target dependent variable prediction model for assisting in the diagnosis of the target dependent variable.

[0070] It should be noted that the target dependent variable prediction model refers to one or more models that meet the prediction needs of the target user to predict the first dependent variable and help the target user make a diagnostic decision.

[0071] Specifically, a horizontal comparison is made among the to-be-determined dependent variable prediction models to obtain a model comparison result. Then, a directional matching is performed on the to-be-determined dependent variable prediction models to obtain a model matching result. Based on the model comparison result and the model matching result, the target dependent variable prediction model and the model ranking of the target dependent variable prediction model are determined. Then, the target dependent variable prediction model is pushed to the target user according to the model ranking to assist in the diagnosis of the target dependent variable.

[0072] In the embodiment of the present invention, by obtaining the target dependent variable selected by the target user, several first independent variable groups corresponding to the target dependent variable are determined. Then, based on the preset training algorithm and each first independent variable group, the initial dependent variable prediction model is iteratively trained to obtain several to-be-determined dependent variable prediction models corresponding to the target dependent variable. Furthermore, the to-be-determined dependent variable prediction models are sorted by accuracy to determine the target dependent variable prediction model for assisting in the diagnosis of the target dependent variable, thereby solving problems such as scattered data, large data processing workload, difficulty in transforming text into results, and difficulty in integrating AI into clinical decision-making in the medical industry, and improving the convenience of medical treatment.

[0073] Based on this, an embodiment of the present application provides a method for assisting traditional Chinese medicine diagnosis. Referring to Figure 2 , Figure 2 which is a schematic flowchart provided for the second embodiment of the method for assisting traditional Chinese medicine diagnosis of the present application.

[0074] In a feasible implementation manner, the determining of a plurality of first independent variable groups corresponding to the target dependent variable includes:

[0075] Step S21: Query the target dependent variable with a preset relationship database to determine a plurality of second independent variables corresponding to the target dependent variable;

[0076] It should be noted that for each data variable index, some are natural independent variables (such as age, gender, getting sick, etc.), some are natural dependent variables (such as improvement, cure, death, etc.), and some are independent variables in some cases and become dependent variables in some other cases (such as taking a certain medicine after getting a certain disease, the medicine is a dependent variable; but after taking the medicine, a certain result is produced, then the medicine is an independent variable). Therefore, it is necessary to distinguish and judge in what situation an index is an independent variable or a dependent variable. The preset relationship database refers to a database that is pre-set and stores rules and related information for judging the causal relationship between independent variables and dependent variables, including natural independent variables (such as age, gender, medical history, etc. always as independent variables), natural dependent variables (such as treatment effect, disease improvement, etc. always as dependent variables), prior experience (such as medical knowledge base, etc.), preset judgment regular expressions (used to judge whether it is an independent variable or a dependent variable according to the type of variable and specific situation), and preset relationship rules (such as event A occurs leading to event B occurring, A is an independent variable, and B is a dependent variable), etc. There is no limitation here and it can be set according to the actual situation.

[0077] Further, it should be noted that the second independent variable refers to the independent variable index that is successfully matched after querying the target dependent variable with the preset relationship database.

[0078] Specifically, query the target dependent variable with the preset relationship database to determine a plurality of second independent variables corresponding to the target dependent variable. Among them, query can be performed through semantic analysis and word segmentation matching and other methods to improve the matching degree and accuracy of the query.

[0079] Step S22: Classify each of the second independent variables to obtain a plurality of second independent variable groups;

[0080] It should be noted that the second independent variable group refers to the group of second independent variables obtained after classifying different categories of second independent variables.

[0081] Specifically, iterative training is performed according to preset independent variable samples and an automatic determination algorithm to obtain an independent variable classifier, and then each of the second independent variables is classified by the independent variable classifier to obtain a plurality of second independent variable groups.

[0082] Step S23: determining a plurality of first independent variable groups corresponding to the target dependent variable based on each of the second independent variable groups.

[0083] Specifically, the value rate corresponding to each second independent variable is obtained, and then the value rates of the second independent variables in each second independent variable group are sorted, and the second independent variables in each group that do not meet the preset value rate range are eliminated, so that the eliminated second independent variable group is used as the first independent variable group.

[0084] This embodiment determines a number of second independent variables corresponding to the target dependent variable by querying the target dependent variable with a preset relational database, and then classifies each of the second independent variables to obtain a number of second independent variable groups, thereby determining a number of first independent variable groups corresponding to the target dependent variable based on each of the second independent variable groups, which helps to quickly clarify the relationship between the target dependent variable and other variables, thereby effectively improving the comprehensiveness of independent variable coverage and the prediction accuracy of subsequent models.

[0085] In a feasible implementation manner, determining a plurality of first independent variable groups corresponding to the target dependent variable based on each of the second independent variable groups includes:

[0086] Step S31, obtaining the value rate corresponding to each of the second independent variables;

[0087] It should be noted that the value rate refers to the proportion of data that is valid, available or valuable for the second independent variable. For example, 50% of the patients did not talk about their daily urination, so the value rate of the second independent variable (i.e., daily urination) is 50%; 30% of the patients did not talk about their sleep conditions, so the value rate of the second independent variable (i.e., sleep conditions) is 70%; and all patients described the condition of their tongue coating, so the value rate of the second independent variable (i.e., tongue coating) is 100%.

[0088] Specifically, statistical analysis is performed on the patient data corresponding to each of the second independent variables to obtain the value rate corresponding to each of the second independent variables.

[0089] Step S32, sorting the value rates of the second independent variables in each of the second independent variable groups, and eliminating the second independent variables in each group that do not meet the preset value rate range;

[0090] It should be noted that the preset value rate range refers to the range corresponding to the value rates of different standards. For example, when the value rate is 0% - 65%, the corresponding second independent variable is excluded. There is no limitation here and it can be restricted according to the actual situation.

[0091] Specifically, sort the value rates of the second independent variables in each of the second independent variable groups. For example, the higher the value rate, the more forward the sorting. Then exclude the second independent variables in each group that do not meet the preset value rate range. For example, when the sorting of the value rates of the second independent variables in one of the second independent variable groups is: 1. Second independent variable B, value rate 85%; 2. Second independent variable C, value rate 78%; 3. Second independent variable A, value rate 40%, and the preset value rate range is 65% - 100%, then the second independent variable A is excluded.

[0092] In addition, it is also possible to calculate the total number of independent variables in each of the second independent variable groups after exclusion, and determine several second independent variable groups that meet the preset total number of independent variables range. Here, the total number of independent variables refers to the total number of independent variables in a group of second independent variable groups, and the preset total number of independent variables range refers to the range of the preset total number of independent variables that meet the requirements. For example,

[0093] Step S33, use the second independent variable group after exclusion as the first independent variable group.

[0094] In this embodiment, by obtaining the value rates corresponding to each of the second independent variables, then sorting the value rates of the second independent variables in each of the second independent variable groups, and excluding the second independent variables in each group that do not meet the preset value rate range, so as to use the second independent variable group after exclusion as the first independent variable group, thereby improving the quality of the data and constructing a more reliable data model, which helps to improve the prediction accuracy of the model and enhance the credibility of the prediction.

[0095] In a feasible implementation manner, iteratively training the initial dependent variable prediction model based on the preset training algorithm and each of the first independent variable groups to obtain several pending dependent variable prediction models corresponding to the target dependent variable, including:

[0096] Step S41, associatively combine each of the first independent variable groups to obtain several groups of training samples;

[0097] It should be noted that the training samples refer to the training samples used for model training, including the training samples and the sample labels corresponding to the training samples.

[0098] Specifically, associatively combine each of the first independent variable groups. For example, if model training is performed for all combination forms, the total number of combinations after associatively combining each of the first independent variable groups is: Cn 2 +Cn3 … + Cn n = 2 n - Cn 0 - Cn 1 = 2 n -n - 1, obtaining several groups of training samples.

[0099] Step S42, input each group of the training samples into initial dependent variable prediction models applying different preset training algorithms for iterative training, obtaining several undetermined dependent variable prediction models corresponding to the target dependent variable.

[0100] Specifically, input each group of the training samples into initial dependent variable prediction models applying different preset training algorithms for iterative training, obtaining several undetermined dependent variable prediction models corresponding to the target dependent variable. In one embodiment, the training process of one of the undetermined dependent variable prediction models is as follows: perform data preprocessing on each group of the training samples, such as data cleaning, normalization, etc., so as to input each group of the training samples and the sample labels corresponding to each group of the training samples into the initial dependent variable prediction model applying the SVM algorithm, obtaining the predicted values output by the initial dependent variable prediction model, and then based on the predicted values and the sample labels, calculating the model loss value using a loss function. In this embodiment, the loss function can be set according to actual requirements and is not specifically limited here. After calculating the model loss value, this training process ends, and then use the error backpropagation algorithm to update the model parameters in the initial dependent variable prediction model, and then perform the next training. During the training process, determine whether the updated initial dependent variable prediction model meets the preset training end conditions. If it meets, use the updated initial dependent variable prediction model as the undetermined dependent variable prediction model. If it does not meet, continue to train the model, where the preset training end conditions include loss convergence and reaching the maximum iteration number threshold, etc.

[0101] In this embodiment, by associatively combining each of the first independent variable groups, several groups of training samples are obtained. Furthermore, each group of the training samples is respectively input into initial dependent variable prediction models applying different preset training algorithms for iterative training, obtaining several undetermined dependent variable prediction models corresponding to the target dependent variable, thereby constructing multiple different prediction models, increasing the diversity of the models, helping to discover data laws from different perspectives, and at the same time improving the generalization ability of the models, further improving the efficiency of medical research, saving time and computing resources, and solving problems such as data dispersion, large data processing workload, difficulty in transforming text into results, and difficulty in integrating AI into clinical decision-making in the medical industry.

[0102] Based on this, an embodiment of the present application provides a traditional Chinese medicine medical diagnosis assistance method, referring to Figure 3 , Figure 3This is a schematic flowchart provided for the third embodiment of the traditional Chinese medicine diagnosis assistance method in this application.

[0103] In a feasible implementation manner, the step of sorting the accuracy of each of the to-be-determined dependent variable prediction models to determine a target dependent variable prediction model for diagnosing assistance for the target dependent variable includes:

[0104] Step S51: Horizontally compare each of the to-be-determined dependent variable prediction models to obtain a model comparison result;

[0105] It should be noted that the model comparison result refers to the result obtained after horizontally comparing each of the to-be-determined dependent variable prediction models, so as to determine the evaluation ranking result of each of the to-be-determined dependent variable prediction models in terms of performance such as accuracy and stability.

[0106] Specifically, obtain the stability evaluation result, accuracy rate, precision rate, and recall rate of each of the to-be-determined dependent variable prediction models, and then calculate the accuracy rate, the precision rate, and the recall rate to determine the accuracy evaluation result of the to-be-determined dependent variable prediction model, so as to obtain the model comparison result based on the stability evaluation result and the accuracy evaluation result.

[0107] Step S52: Align the directions of each of the to-be-determined dependent variable prediction models to obtain a model alignment result;

[0108] It should be noted that direction alignment is used to characterize the correlation or consistency between two or more variables. Especially in time series data or ordered data, it refers to whether the directions of variable changes are the same. In the present invention, the prediction result and the prediction variable of the to-be-determined dependent variable prediction model are judged by the way of direction alignment. The model alignment result refers to the result obtained after aligning the directions of each of the to-be-determined dependent variable prediction models.

[0109] Specifically, align the directions of each of the to-be-determined dependent variable prediction models. Among them, the calculation of direction alignment can be performed by DTW (Dynamic Time Warping). That is, the distance matrix obtained by DTW can be used to judge the degree of direction alignment. There is no limitation here, and it can be set according to the actual situation, so as to determine the model alignment result according to the alignment degree results of each of the to-be-determined dependent variable prediction models.

[0110] Additionally, based on the model fitting results, the to-be-determined dependent variable prediction models with directional convergence (converging with human cognition direction, having high interpretability, facilitating human evaluation and application promotion) can be retained, and the to-be-determined dependent variable prediction models with chaotic directionality (different from existing human cognition and with unclear direction, some being positively correlated with a certain view but many being negatively correlated with it) can be excluded, thereby increasing the reliability and practicality of the models, reducing the misuse risk of unreliable models, and reducing the cognitive conflict and misuse risk, and promoting the integration and expansion of knowledge.

[0111] Step S53: Based on the model comparison result and the model fitting result, determine the target dependent variable prediction model and the model ranking of the target dependent variable prediction model.

[0112] It should be noted that the model ranking refers to the comprehensive ranking after multi-faceted evaluation of the target dependent variable prediction model.

[0113] Specifically, based on the model comparison result and the model fitting result, determine the target dependent variable prediction model and the model ranking of the target dependent variable prediction model. For example, directly use the superposition calculation method of the model comparison result and the model fitting result to determine that the top 40% of the to-be-determined dependent variable prediction models are the target dependent variable prediction models, and obtain the model ranking according to the calculation result. Another example is to perform weighted calculation on the model comparison result and the model fitting result to determine that the top 50% of the to-be-determined dependent variable prediction models are the target dependent variable prediction models, and obtain the model ranking according to the calculation result. There is no limitation here and it can be set according to the actual situation.

[0114] Step S54: Push the target dependent variable prediction model to the target user according to the model ranking to assist in diagnosing the target dependent variable.

[0115] Specifically, push the target dependent variable prediction model to the target user according to the model ranking for the target user to assist in diagnosing the target dependent variable based on the predicted value output by the target dependent variable prediction model. Among them, it can be pushed through ways such as web pages, APPs, and mini-programs, and there is no limitation here.

[0116] In this embodiment, by making a horizontal comparison of each of the to-be-determined dependent variable prediction models, a model comparison result is obtained. Furthermore, by making a directional match for each of the to-be-determined dependent variable prediction models, a model matching result is obtained. Thus, based on the model comparison result and the model matching result, a target dependent variable prediction model and the model ranking of the target dependent variable prediction model are determined. Then, the target dependent variable prediction model is pushed to the target user according to the model ranking to assist in diagnosing the target dependent variable, thereby ensuring the efficiency and accuracy of the model, improving the diagnostic accuracy, reducing the possibility of misdiagnosis and missed diagnosis, and further realizing personalized medicine and precision treatment, thus improving the efficiency of medical decision-making and promoting the development of the medical industry.

[0117] In a feasible implementation manner, the making a horizontal comparison of each of the to-be-determined dependent variable prediction models to obtain a model comparison result includes:

[0118] Step S61, obtaining the stability evaluation result, accuracy rate, precision rate, and recall rate of each of the to-be-determined dependent variable prediction models;

[0119] It should be noted that the stability evaluation result refers to the evaluation result of the model stability, the accuracy rate refers to the proportion of the number of correctly predicted samples to the total number of samples, the precision rate refers to the proportion of true positives (TP) to the number of samples predicted as positive (TP + FP), and the recall rate refers to the proportion of true positives (TP) to the number of samples actually being positive (TP + FN).

[0120] Specifically, the stability evaluation result, accuracy rate, precision rate, and recall rate of each of the to-be-determined dependent variable prediction models are obtained. Among them, the stability evaluation result can be obtained by using cross-validation (such as k-fold cross-validation) to effectively reduce the influence of overfitting and provide an evaluation of the model stability. The accuracy rate, precision rate, and recall rate can be calculated through the predicted values output by the to-be-determined dependent variable prediction models, and there is no limitation here.

[0121] Step S62, calculating the accuracy rate, precision rate, and recall rate to determine the accuracy evaluation result of the to-be-determined dependent variable prediction model;

[0122] It should be noted that the accuracy evaluation result refers to the evaluation result of the model accuracy.

[0123] Specifically, the accuracy, precision, and recall are comprehensively calculated to determine the accuracy evaluation result of the to-be-determined dependent variable prediction model. For example, the F1 score (the harmonic mean of precision and recall) and the mean squared error (MSE, the average of the squared differences between the predicted values and the actual values) are calculated, etc., to determine the accuracy evaluation result of the to-be-determined dependent variable prediction model. There is no limitation here, and it can be restricted according to the actual situation.

[0124] Step S63: Obtain a model comparison result based on the stability evaluation result and the accuracy evaluation result.

[0125] Specifically, in one embodiment, the stability evaluation result and the accuracy evaluation result are weighted and assigned values to obtain a model comparison result. In addition, the stability evaluation result and the accuracy evaluation result can also be comprehensively evaluated by means of manual determination, etc., to obtain a model comparison result. There is no limitation here.

[0126] In addition, the stability evaluation result and the accuracy evaluation result can be used to reveal the performance of the model in different aspects, such as false positives (low precision), false negatives (low recall), or stability problems. At the same time, the relative performance of the models can also be compared through the stability evaluation result and the accuracy evaluation result, helping the target user to determine which model is more excellent and has stronger robustness in a specific scenario, increasing the application scenarios of the model. In addition, by recording the evaluation data of each model, the continuous optimization and iteration of the model can be promoted, further improving the accuracy of the model.

[0127] In this embodiment, by obtaining the stability evaluation result, accuracy, precision, and recall of each to-be-determined dependent variable prediction model, and then calculating the accuracy, precision, and recall, the accuracy evaluation result of the to-be-determined dependent variable prediction model is determined. Then, based on the stability evaluation result and the accuracy evaluation result, a model comparison result is obtained. Furthermore, the performance of the model is comprehensively evaluated through multiple indicators, thereby improving the trust in the model and ensuring the reliability of the model.

[0128] In a feasible implementation manner, after ranking the accuracies of the to-be-determined dependent variable prediction models to determine the target dependent variable prediction model for diagnostic assistance for the target dependent variable, it further includes:

[0129] Step S71: Obtain the user score of the target dependent variable prediction model.

[0130] It should be noted that the user score refers to the total score obtained after the user scores the indicators such as the accuracy and stability of the target dependent variable prediction model. Among them, the scores of multiple indicators can be directly added to obtain the user score, or the user score can be calculated by weighting according to the weights of different indicators, so as to determine the user score of the target dependent variable prediction model through a more accurate and objective scoring calculation rule.

[0131] Specifically, in one embodiment, after detecting that the target user has used the target dependent variable prediction model, a scoring questionnaire is generated and the scoring file is pushed to the target user for the target user to score, and the user score is obtained. The acquisition method is not limited here and can be set according to the actual situation.

[0132] Step S72, if the user score meets the preset recommended score range, add the target dependent variable prediction model and the usage instructions corresponding to the target dependent variable prediction model to the recommended model library;

[0133] It should be noted that the preset recommended score range refers to the score range set in advance. For example, a score of 1-59 is not recommended, a score of 60-79 is pending, and a score of 80-100 is recommended. The specific score range can be set according to the actual situation and is not limited here. The usage instructions refer to the usage rules of the target dependent variable prediction model and model-related information, etc., including independent variable indicators, the first dependent variable, and training algorithms, etc.

[0134] Furthermore, it should be noted that the recommended model library refers to a library that stores target dependent variable prediction models that can be recommended after user scoring, so as to build a technical platform within medical institutions with (traditional Chinese medicine) data, perform automated computing, mining, and processing on the data widely existing in clinical research at any time and anywhere. Traditional Chinese medicine experts only provide directional guidance, determination, and conclusive evaluation. For models that are considered applicable after evaluation, they are directly connected to the medical system process for daily application and scientific research application, etc.

[0135] Specifically, if the user score meets the preset recommended score range, the target dependent variable prediction model and the usage instructions corresponding to the target dependent variable prediction model are associated and combined, and then added to the recommended model library.

[0136] In addition, if the user score does not meet the preset recommended score range, the first independent variable group, the target dependent variable, and the training algorithm corresponding to the target dependent variable prediction model are pushed to the medical expert for the medical expert to confirm whether the model training of this combination needs to be carried out subsequently, so as to save training costs and reduce resource waste.

[0137] Step S73: Push the target dependent variable prediction model to medical experts for model evaluation by the medical experts.

[0138] Specifically, after pushing the target dependent variable prediction model to medical experts for model evaluation by the medical experts, it is necessary to confirm whether to still retain the target dependent variable prediction model in the recommended model library, or confirm whether to perform further iterative updates on the target dependent variable prediction model, etc., so as to put forward improvement suggestions according to actual clinical needs and model performance, and then optimize the model structure and parameters to improve the performance of the model. This is not limited here.

[0139] In this embodiment, by obtaining the user score of the target dependent variable prediction model, and if the user score meets the preset recommended score range, the target dependent variable prediction model and the usage instructions corresponding to the target dependent variable prediction model are added to the recommended model library, so as to push the target dependent variable prediction model to medical experts for model evaluation by the medical experts, thereby improving the accuracy, generalization ability, reliability and applicability of the model, expanding the application scope of the model, promoting medical research, and improving the overall quality of medical services.

[0140] In a feasible implementation manner, before iteratively training the initial dependent variable prediction model based on the preset training algorithm and each of the first independent variable groups to obtain a number of pending dependent variable prediction models corresponding to the target dependent variable, it further includes:

[0141] Step S81: Associatively combine the target dependent variable with each of the first independent variables to obtain variable combinations, and determine whether there are identical variable combinations;

[0142] It should be noted that a variable combination refers to the combination of the target dependent variable and each of the first independent variables. For example, if the target dependent variable is the efficacy of drug xx, and the first independent variables are administration time, patient age, and patient gender, the variable combination is the efficacy of drug xx, administration time, patient age, and patient gender.

[0143] Specifically, associatively combine the target dependent variable with each of the first independent variables to obtain variable combinations, and then query and match the variable combinations with a preset combination database to determine whether there are identical variable combinations. The preset combination database refers to a variable combination database that is pre-set and stores variable combinations for which iterative training has been performed according to a certain variable combination to obtain a target dependent variable prediction model, so as to determine whether the system has performed model training according to the current variable combination based on the preset combination database, thereby reducing resource waste.

[0144] Step S82: If so, use the target dependent variable prediction model corresponding to the variable combination as the target dependent variable prediction model for the target dependent variable to assist in diagnosis.

[0145] Specifically, if there is the same variable combination, determine the target dependent variable prediction model corresponding to the variable combination, and use the target dependent variable prediction model corresponding to the variable combination as the target dependent variable prediction model for the target dependent variable to assist in diagnosis, thereby saving the subsequent model training process, enabling the target user to directly perform diagnosis assistance based on the existing model, saving training resources and time, and improving the medical treatment efficiency.

[0146] In this embodiment, by associating and combining the target dependent variable with each of the first independent variables to obtain variable combinations, and determining whether there are the same variable combinations, and if so, using the target dependent variable prediction model corresponding to the variable combination as the target dependent variable prediction model for the target dependent variable to assist in diagnosis, so that the target user can directly perform diagnosis assistance based on the existing model that is the same as the user's needs, saving the resources and time of model training, and improving the medical treatment efficiency.

[0147] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The order of execution of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0148] This application also provides a traditional Chinese medicine medical diagnosis assistance device. Please refer to Figure 4 , the traditional Chinese medicine medical diagnosis assistance device includes:

[0149] An acquisition module 41, configured to acquire a target dependent variable selected by a target user;

[0150] A determination module 42, configured to determine a plurality of first independent variable groups corresponding to the target dependent variable;

[0151] A training module 43, configured to iteratively train an initial dependent variable prediction model based on a preset training algorithm and each of the first independent variable groups to obtain a plurality of pending dependent variable prediction models corresponding to the target dependent variable;

[0152] A sorting module 44, configured to sort the accuracy of each of the pending dependent variable prediction models to determine a target dependent variable prediction model to assist in diagnosis for the target dependent variable.

[0153] The traditional Chinese medicine medical diagnosis assistance device is also used for:

[0154] Query the preset relational database with the target dependent variable to determine a number of second independent variables corresponding to the target dependent variable;

[0155] Classify each of the second independent variables to obtain a number of second independent variable groups;

[0156] Based on each of the second independent variable groups, determine a number of first independent variable groups corresponding to the target dependent variable.

[0157] The traditional Chinese medicine medical diagnosis assistance device is also used for:

[0158] Obtain the value ratio corresponding to each of the second independent variables;

[0159] Sort the value ratios of the second independent variables in each of the second independent variable groups, and eliminate the second independent variables that do not meet the preset value ratio range in each group;

[0160] Use the eliminated second independent variable groups as the first independent variable groups.

[0161] The traditional Chinese medicine medical diagnosis assistance device is also used for:

[0162] Associate and combine each of the first independent variable groups to obtain several groups of training samples;

[0163] Input each group of training samples into the initial dependent variable prediction models applying different preset training algorithms for iterative training to obtain several dependent variable prediction models to be determined corresponding to the target dependent variable.

[0164] The traditional Chinese medicine medical diagnosis assistance device is also used for:

[0165] Conduct horizontal comparison on each of the dependent variable prediction models to be determined to obtain a model comparison result;

[0166] Conduct directional matching on each of the dependent variable prediction models to be determined to obtain a model matching result;

[0167] Based on the model comparison result and the model matching result, determine the target dependent variable prediction model and the model ranking of the target dependent variable prediction model;

[0168] Push the target dependent variable prediction model to the target user according to the model ranking to assist in diagnosing the target dependent variable.

[0169] The traditional Chinese medicine medical diagnosis assistance device is also used for:

[0170] Obtain the stability evaluation result, accuracy rate, precision rate, and recall rate of each of the dependent variable prediction models to be determined;

[0171] Calculate the accuracy rate, precision rate, and recall rate to determine the accuracy evaluation result of the to-be-determined dependent variable prediction model;

[0172] Based on the stability evaluation result and the accuracy evaluation result, obtain the model comparison result.

[0173] The traditional Chinese medicine medical diagnosis assistance device is also used for:

[0174] Obtain the user score of the target dependent variable prediction model;

[0175] If the user score meets the preset recommended score range, add the target dependent variable prediction model and the usage instructions corresponding to the target dependent variable prediction model to the recommended model library;

[0176] Push the target dependent variable prediction model to medical experts for model evaluation by the medical experts.

[0177] Associate and combine the target dependent variable with each of the first independent variables to obtain variable combinations, and determine whether there are identical variable combinations;

[0178] If so, use the target dependent variable prediction model corresponding to the variable combination as the target dependent variable prediction model corresponding to the target dependent variable for diagnostic assistance for the target dependent variable.

[0179] The traditional Chinese medicine medical diagnosis assistance device provided by the present application adopts the traditional Chinese medicine medical diagnosis assistance method in the above embodiment, and can solve the technical problems in the background art. Compared with the prior art, the beneficial effects of the traditional Chinese medicine medical diagnosis assistance device provided by the present application are the same as those of the traditional Chinese medicine medical diagnosis assistance method provided by the above embodiment, and other technical features in the traditional Chinese medicine medical diagnosis assistance device are the same as the features disclosed in the above embodiment method, and will not be elaborated here.

[0180] The present application provides a traditional Chinese medicine medical diagnosis assistance device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the traditional Chinese medicine medical diagnosis assistance method in the first embodiment above.

[0181] Next, refer to Figure 5, which shows a schematic structural diagram of a traditional Chinese medicine diagnosis assistance device suitable for implementing the embodiments of the present application. The traditional Chinese medicine diagnosis assistance device in the embodiments of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The shown traditional Chinese medicine diagnosis assistance device is merely an example and should not impose any limitations on the functions and usage scope of the embodiments of the present application.

[0182] As Figure 5 shown, the traditional Chinese medicine diagnosis assistance device may include a processing device 1001 (such as a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM: Read Only Memory) 1002 or the program loaded from the storage device 1003 into the random access memory (RAM: Random Access Memory) 1004. In the RAM 1004, various programs and data required for the operation of the traditional Chinese medicine diagnosis assistance device are also stored. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other through a bus 1005. The input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems may be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the traditional Chinese medicine diagnosis assistance device to communicate with other devices wirelessly or wireline to exchange data. Although the figure shows a traditional Chinese medicine diagnosis assistance device with various systems, it should be understood that it is not required to implement or have all the shown systems. More or fewer systems may be implemented or had alternatively.

[0183] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program codes for executing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by a processing device 1001, the above-mentioned functions defined in the methods of the embodiments disclosed in the present application are executed.

[0184] The traditional Chinese medicine medical diagnosis assistance device provided by the present application adopts the traditional Chinese medicine medical diagnosis assistance method in the above embodiment, and can solve the technical problems in the background art. Compared with the prior art, the beneficial effects of the traditional Chinese medicine medical diagnosis assistance device provided by the present application are the same as those of the traditional Chinese medicine medical diagnosis assistance method provided in the above embodiment, and other technical features in the traditional Chinese medicine medical diagnosis assistance device are the same as the features disclosed in the method of the previous embodiment, and will not be elaborated here.

[0185] It should be understood that each part disclosed in the present application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0186] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, and all of them should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0187] The present application provides a computer-readable storage medium, on which there are computer-readable program instructions (i.e., computer programs), and the computer-readable program instructions are used to execute the traditional Chinese medicine medical diagnosis assistance method in the above embodiment.

[0188] The computer-readable storage medium provided by this application can be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) or flash memory, optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this embodiment, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, device, or component. The program code contained on the computer-readable storage medium can be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0189] The above computer-readable storage medium can be included in a traditional Chinese medicine medical diagnosis assistance device; or it can exist separately without being assembled into the traditional Chinese medicine medical diagnosis assistance device.

[0190] The above computer-readable storage medium carries one or more programs. When the above one or more programs are executed by a traditional Chinese medicine medical diagnosis assistance device, the traditional Chinese medicine medical diagnosis assistance device is caused to:

[0191] Obtain the target dependent variable selected by the target user;

[0192] Determine several first independent variable groups corresponding to the target dependent variable;

[0193] Based on a preset training algorithm and each of the first independent variable groups, iteratively train an initial dependent variable prediction model to obtain several pending dependent variable prediction models corresponding to the target dependent variable;

[0194] Sort the accuracies of each of the pending dependent variable prediction models to determine a target dependent variable prediction model for assisting in the diagnosis of the target dependent variable.

[0195] Computer program code for performing the operations of this application can be written in one or more programming languages or combinations thereof. The above-mentioned programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any kind of network, including a local area network (LAN: Local Area Network) or a wide area network (WAN: Wide Area Network), or it can be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0196] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutively represented blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0197] The modules described in the embodiments of this application can be implemented in software or in hardware. Among them, the name of the module does not constitute a limitation on the unit itself in some cases.

[0198] The readable storage medium provided by this application is a computer-readable storage medium. The computer-readable storage medium stores computer-readable program instructions (i.e., computer programs) for performing the above-mentioned traditional Chinese medicine diagnosis assistance method, and can solve the technical problems as in the background art. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by this application are the same as those of the traditional Chinese medicine diagnosis assistance method provided by the above embodiments, and will not be elaborated here.

[0199] An embodiment of the present application provides a computer program product, including a computer program, which when executed by a processor implements the steps of the above-mentioned traditional Chinese medicine diagnosis assistance method.

[0200] The computer program product provided by the present application can solve the technical problems in the background art. Compared with the prior art, the beneficial effects of the computer program product provided by the embodiment of the present application are the same as those of the traditional Chinese medicine diagnosis assistance method provided by the above embodiment, and will not be elaborated here.

[0201] The above are only partial embodiments of the present application, and do not limit the patent scope of the present application. Any equivalent structural transformation made by using the content of the specification and drawings of the present application under the technical concept of the present application, or direct / indirect application in other related technical fields, is included in the patent protection scope of the present application.

Claims

1. A Chinese medicine diagnosis auxiliary method, characterized in that: include: Obtain the target dependent variable selected by the target user; Determine a plurality of first independent variable groups corresponding to the target dependent variable; Based on a preset training algorithm and each of the first independent variable groups, iteratively train the initial dependent variable prediction model to obtain a plurality of pending dependent variable prediction models corresponding to the target dependent variable; The accuracy of each of the pending dependent variable prediction models is ranked, and the target dependent variable prediction model is determined to assist in diagnosis of the target dependent variable.

2. The TCM diagnosis auxiliary method according to claim 1, characterized in that: The determining of a plurality of first independent variable groups corresponding to the target dependent variable comprises: Querying the target dependent variable with a preset relational database to determine a plurality of second independent variables corresponding to the target dependent variable; Classifying each of the second independent variables to obtain a plurality of second independent variable groups; Based on each of the second independent variable groups, a number of first independent variable groups corresponding to the target dependent variable are determined.

3. The TCM diagnosis auxiliary method according to claim 2, characterized in that: The determining, based on each of the second independent variable groups, a plurality of first independent variable groups corresponding to the target dependent variable comprises: Obtaining the value rate corresponding to each of the second independent variables; Sort the value rates of the second independent variables in each of the second independent variable groups, and remove the second independent variables in each group that do not meet the preset value rate range; The eliminated second independent variable group is used as the first independent variable group.

4. The TCM diagnosis auxiliary method according to claim 1, characterized in that: The initial dependent variable prediction model is iteratively trained based on the preset training algorithm and each of the first independent variable groups to obtain a plurality of pending dependent variable prediction models corresponding to the target dependent variable, including: Associating and combining the first independent variable groups to obtain a plurality of training sample groups; The groups of training samples are respectively input into initial dependent variable prediction models using different preset training algorithms for iterative training to obtain a number of pending dependent variable prediction models corresponding to the target dependent variables.

5. The TCM diagnosis auxiliary method according to claim 1, characterized in that: The step of ranking the accuracy of the prediction models of the undetermined dependent variables and determining the prediction model of the target dependent variable to assist in diagnosis of the target dependent variable includes: Performing a horizontal comparison on the prediction models of the undetermined dependent variables to obtain a model comparison result; Directionally matching the prediction models of the undetermined dependent variables to obtain model matching results; Determine the target dependent variable prediction model and the model ranking of the target dependent variable prediction model based on the model comparison result and the model matching result; The target dependent variable prediction model is pushed to the target user according to the model ranking to assist in diagnosis of the target dependent variable.

6. The TCM diagnosis auxiliary method according to claim 5, characterized in that: The horizontal comparison of the prediction models of the undetermined dependent variables to obtain the model comparison results includes: Obtaining the stability evaluation results, accuracy, precision and recall of each of the pending dependent variable prediction models; Calculating the accuracy rate, the precision rate, and the recall rate to determine the accuracy evaluation result of the prediction model of the undetermined dependent variable; Based on the stability evaluation result and the accuracy evaluation result, a model comparison result is obtained.

7. The TCM diagnosis auxiliary method according to claim 1, characterized in that: After ranking the accuracy of the prediction models of the undetermined dependent variables and determining the prediction model of the target dependent variable to assist in diagnosis of the target dependent variable, the method further includes: Obtaining user ratings of the target dependent variable prediction model; If the user rating meets the preset recommendation score range, the target dependent variable prediction model and the usage instructions corresponding to the target dependent variable prediction model are added to the recommendation model library; The target dependent variable prediction model is pushed to a medical expert for the medical expert to perform model evaluation.

8. The TCM diagnosis auxiliary method according to claim 1, characterized in that: Before the iterative training of the initial dependent variable prediction model based on the preset training algorithm and each of the first independent variable groups to obtain a plurality of pending dependent variable prediction models corresponding to the target dependent variable, the method further includes: Associating and combining the target dependent variable with each of the first independent variables to obtain a variable combination, and determining whether there is an identical variable combination; If so, the target dependent variable prediction model corresponding to the variable combination is used as the target dependent variable prediction model corresponding to the target dependent variable to assist in diagnosis of the target dependent variable.

9. A Chinese medicine diagnosis auxiliary device, characterized in that: The TCM medical diagnosis auxiliary device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the TCM medical diagnosis auxiliary method as described in any one of claims 1 to 8.

10. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the TCM medical diagnosis auxiliary method as described in any one of claims 1 to 8 are implemented.

Images