A method and system for intelligent analysis of physical therapy data

Abstract

The present invention discloses a method and system for intelligent analysis of physiotherapy data, and relates to the field of medical health data analysis. A method and system for intelligent analysis of physiotherapy data, including: a data acquisition module, an intelligent analysis module, a physiotherapy plan generation module, a monitoring feedback module, an effect evaluation module and a visual interface module. The present invention obtains time series data of pulse, blood pressure and body temperature, including static analysis data in a stationary state and real-time physiotherapy data during physiotherapy, obtains static analysis results through a static data analysis model, obtains real-time physiotherapy data analysis results through a real-time physiotherapy data analysis model, the static analysis results are used to evaluate the health status of the user, and the real-time physiotherapy data analysis results are used to evaluate the effect of the physiotherapy plan, and provides the user with a more accurate and objective physiotherapy plan through the user's health status and the effect of the physiotherapy plan.

Description

Technical Field

[0001] The present invention relates to the field of medical health data analysis, and in particular to a method and system for intelligent analysis of physical therapy data. Background Art

[0002] In the field of medical health, physiotherapy, as an important rehabilitation treatment method, is widely used in the rehabilitation treatment of various sports injuries, neurological dysfunction and other diseases. With the advancement of science and technology, traditional manual physiotherapy has gradually evolved into a trend of automated and precise treatment using physiotherapy equipment. Among them, pneumatic physiotherapy equipment has the advantages of safety and controllability compared to other physiotherapy equipment. Pneumatic physiotherapy equipment can provide more standardized and personalized treatment plans by precisely controlling air pressure and motion trajectory, thereby improving treatment effects and patients' treatment experience.

[0003] When conducting physical therapy, data collection and analysis are essential for evaluating treatment effects, adjusting treatment plans, and personalizing patient care. However, there are still some challenges in physical therapy data processing and effect analysis. Most traditional physical therapy data analysis methods are limited to basic data statistics and display, and fail to fully utilize the potential value behind the data. Although this method can provide a certain degree of feedback on treatment effects, it lacks in-depth data mining and personalized analysis, and cannot fully utilize the potential of pneumatic physical therapy equipment. On the other hand, the development of modern medical informatization has led to an increase in the demand for data processing and analysis capabilities during physical therapy. Medical practitioners hope to improve the personalization of treatment effects and the accuracy of objective evaluation through more intelligent data analysis methods. Summary of the invention

[0004] A method for intelligent analysis of physical therapy data, comprising:

[0005] S1. Collect relevant data of the user, including gender, age, height, weight and disease type, as basic physical data; time series data of pulse, blood pressure and body temperature in a static state, as static analysis data; time series data of pulse, blood pressure, body temperature and physical therapy methods during physical therapy, as real-time physical therapy data;

[0006] S2. Preprocess the user's relevant data, extract features from the preprocessed static analysis data and real-time physical therapy data, and organize the extracted features and basic body data into physical therapy object data;

[0007] S3. The therapy object data is sent to the static data analysis model for analysis to obtain static data analysis results; the static data analysis results and the therapy object data are sent to the real-time therapy data analysis model for analysis to obtain real-time therapy data analysis results;

[0008] S4. Matching a standard physiotherapy plan in the physiotherapy plan library according to the static data analysis results and the real-time physiotherapy data analysis results, generating physiotherapy suggestions, and outputting the standard physiotherapy plan and physiotherapy suggestions;

[0009] S5. Collect relevant data of the user according to the set monitoring time, obtain static data analysis results and real-time physical therapy data analysis results of the relevant data of the user collected at that time, and use them as monitoring feedback data, adjust the physical therapy plan according to the monitoring feedback data, and generate feedback physical therapy suggestions;

[0010] S6. Perform a physiotherapy effect evaluation based on the static data analysis results of the user's relevant data collected each time, and generate a physiotherapy effect evaluation table;

[0011] S7. Visualize the basic body data, physical therapy plan, physical therapy suggestions and physical therapy effect evaluation form, and output the visualization results.

[0012] As a preferred technical solution of the present invention, the static data analysis model includes a data processing layer and a static data analysis layer;

[0013] The data processing layer is used to pre-process the therapy object data to obtain static data to be analyzed;

[0014] The static data analysis layer is used to analyze the static data to be analyzed and output the static data analysis results;

[0015] Build a static data analysis model. The specific steps are:

[0016] A1. Obtain a number of physical therapy object data, preprocess the physical therapy object data and mark the static analysis result labels, form all the physical therapy object data that have been preprocessed and marked with static analysis result labels into a static analysis data set, and divide the static analysis data set into a static analysis training set and a static analysis test set;

[0017] A2. Send the static analysis training set to the initial static data analysis model based on the recurrent neural network for training, and use the static analysis result label as the target to train the optimized static data analysis model;

[0018] A3. Send the static analysis test set to the optimized static data analysis model for accuracy evaluation, and use the static analysis result label as the target to obtain the test accuracy. If the test accuracy is within the error standard range, the optimized static data analysis model will be used as the static data analysis model. If the test accuracy exceeds the error standard range, adjust the hyperparameters of the optimized static data analysis model and repeat steps A2-A3.

[0019] As a preferred technical solution of the present invention, the real-time physical therapy data analysis model includes a data processing layer and a real-time data analysis layer;

[0020] The data processing layer is used to reconstruct the therapy object data according to the static data analysis results to obtain the reconstructed therapy object data, and preprocess the data to obtain the real-time therapy data to be analyzed;

[0021] The real-time data analysis layer is used to analyze the real-time physical therapy data to be analyzed and output the real-time physical therapy data analysis results;

[0022] Build a real-time physical therapy data analysis model. The specific steps are as follows:

[0023] B1. Obtain some physiotherapy object data and static data analysis results, reconstruct the physiotherapy object data according to the static data analysis results to obtain reconstructed physiotherapy object data, preprocess the reconstructed physiotherapy object data and mark the real-time physiotherapy data analysis result labels, form all the reconstructed physiotherapy object data that have been preprocessed and marked with real-time physiotherapy data analysis result labels into a real-time physiotherapy analysis data set, and divide the real-time physiotherapy analysis data set into a real-time physiotherapy analysis training set and a real-time physiotherapy analysis test set;

[0024] B2. Send the real-time physical therapy analysis training set to the initial real-time physical therapy data analysis model based on the long short-term memory network for training, and take the real-time physical therapy data analysis result label as the target to train and obtain the optimized real-time physical therapy data analysis model;

[0025] B3. Send the real-time physical therapy analysis test set to the optimized real-time physical therapy data analysis model for accuracy evaluation, and use the real-time physical therapy data analysis result label as the target to obtain the test accuracy; if the test accuracy is within the error standard range, the optimized real-time physical therapy data analysis model will be used as the real-time physical therapy data analysis model; if the test accuracy exceeds the error standard range, adjust the hyperparameters of the optimized real-time physical therapy data analysis model and repeat steps B2-B3.

[0026] As a preferred technical solution of the present invention, the physiotherapy object data is reconstructed according to the static data analysis results to obtain reconstructed physiotherapy object data. The specific method of reconstruction is: replacing the features corresponding to the static analysis data with the static analysis results, and adding weights to the feature data corresponding to the real-time physiotherapy data according to the static analysis results.

[0027] As a preferred technical solution of the present invention, feature extraction is performed on the pre-processed static analysis data and real-time physical therapy data. The specific steps of feature extraction are:

[0028] C1. Extract time domain features from the time series data of pulse, blood pressure, and body temperature in static analysis data and real-time physical therapy data, including mean and variance values;

[0029] C2. Convert the time series data of pulse, blood pressure and body temperature into frequency domain data through Fourier transform, and extract the frequency domain features of the frequency domain data, including power spectrum density and spectrum diagram;

[0030] C3. Obtain the nonlinear characteristics of time series data of pulse, blood pressure, and body temperature, including sample entropy and Lyapunov coefficient.

[0031] As a preferred technical solution of the present invention, a standard physiotherapy plan is matched in a physiotherapy plan library according to static data analysis results and real-time physiotherapy data analysis results. The matching method is: disease labels are matched in the physiotherapy plan library through static data analysis results to obtain disease matching results, physiotherapy plans are matched in the disease matching results through real-time physiotherapy data analysis results, and a matched standard physiotherapy plan is output.

[0032] As a preferred technical solution of the present invention, the physiotherapy plan is adjusted according to the monitoring feedback data, and feedback physiotherapy suggestions are generated. The specific steps of the method are:

[0033] D1. The feedback data is matched again in the physiotherapy program library to obtain the feedback disease matching results and feedback physiotherapy programs;

[0034] D2. Adjust the physical therapy plan through the feedback physical therapy plan, and generate feedback physical therapy suggestions based on the feedback disease matching results and the feedback physical therapy plan.

[0035] As a preferred technical solution of the present invention, it is characterized in that a physiotherapy effect evaluation is performed based on the static data analysis results of the relevant data of the user collected each time, and a physiotherapy effect evaluation table is generated; the physiotherapy effect evaluation is obtained by comparing different static data analysis results, and the physiotherapy effect evaluation table is used to record all physiotherapy plans and physiotherapy effect evaluations of the user.

[0036] A physical therapy data intelligent analysis system, comprising:

[0037] The data collection module is used to collect relevant data of the user, including gender, age, height, weight and disease type, as basic body data; time series data of pulse, blood pressure and body temperature in a static state, as static analysis data; time series data of pulse, blood pressure and body temperature during physical therapy and physical therapy methods, as real-time physical therapy data, pre-process the relevant data of the user, extract features from the pre-processed static analysis data and real-time physical therapy data, and organize the extracted features and basic body data into physical therapy object data;

[0038] The intelligent analysis module is used to send the therapy object data into the static data analysis model for analysis to obtain the static data analysis results; send the static data analysis results and the therapy object data into the real-time therapy data analysis model for analysis to obtain the real-time therapy data analysis results;

[0039] A physiotherapy plan generation module is used to match standard physiotherapy plans in the physiotherapy plan library according to the static data analysis results and the real-time physiotherapy data analysis results, generate physiotherapy suggestions, and output the standard physiotherapy plans and physiotherapy suggestions;

[0040] The monitoring feedback module is used to collect relevant data of the user according to the set monitoring time, obtain the static data analysis results and real-time physical therapy data analysis results of the relevant data of the user collected at this time, and use them as monitoring feedback data, adjust the physical therapy plan according to the monitoring feedback data, and generate feedback physical therapy suggestions;

[0041] The effect evaluation module is used to evaluate the physical therapy effect according to the static data analysis results of the relevant data of the user collected each time, and generate a physical therapy effect evaluation table;

[0042] The visualization interface module is used to visualize basic body data, physiotherapy plans, physiotherapy suggestions and physiotherapy effect evaluation tables, and output visualization results.

[0043] The present invention has the following advantages:

[0044] 1. The present invention obtains time series data of pulse, blood pressure and body temperature, including static analysis data in a stationary state and real-time physical therapy data during physical therapy, obtains static analysis results through a static data analysis model, and obtains real-time physical therapy data analysis results through a real-time physical therapy data analysis model. Based on these easily obtained data, the data of the physical therapy user are analyzed in detail. The static analysis results are used to evaluate the user's health status, and the real-time physical therapy data analysis results are used to evaluate the effect of the physical therapy plan. A more accurate and objective physical therapy plan is provided to the user based on the user's health status and the effect of the physical therapy plan.

[0045] 2. The present invention collects relevant data of the user by setting the monitoring time, performs data analysis, and generates monitoring feedback results. The physical therapy plan is adjusted according to the monitoring feedback results, thereby providing the user with a more personalized physical therapy plan.

[0046] 3. The present invention reconstructs the user's physiotherapy object data through static analysis results to obtain reconstructed physiotherapy data, so that the real-time physiotherapy data analysis model can analyze the physiotherapy object data in combination with the static analysis results, so that the model can fully explore the potential value of the data when the amount of data is small, thereby improving the recognition accuracy of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A schematic diagram of the structure of a physical therapy data intelligent analysis system used in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to enable persons skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0049] Embodiment 1, a method for intelligent analysis of physical therapy data, comprising:

[0050] S1. Collect relevant data of the user, including gender, age, height, weight and disease type, as basic physical data; time series data of pulse, blood pressure and body temperature in a static state, as static analysis data; time series data of pulse, blood pressure, body temperature and physical therapy methods during physical therapy, as real-time physical therapy data;

[0051] The data of pulse, blood pressure and body temperature are obtained through the corresponding collectors. The acquisition of these three types of data is convenient and fast, and the cost of the collector is low. The purpose of collecting time series data is to expand the amount of data to reduce the impact of abnormal data on the analysis.

[0052] The physiotherapy method refers to the location and time of the physiotherapy. The physiotherapy method can be performed manually or using physiotherapy equipment. For example, pneumatic physiotherapy equipment is used to massage different acupoints of the user. In this case, the data of the physiotherapy plan is the acupoint number and the corresponding time of the physiotherapy.

[0053] S2. Preprocess the user's relevant data, extract features from the preprocessed static analysis data and real-time physical therapy data, and organize the extracted features and basic body data into physical therapy object data;

[0054] In step S2, feature extraction is performed on the preprocessed static analysis data and real-time physical therapy data. The specific steps of feature extraction are:

[0055] C1. Extract time domain features from the time series data of pulse, blood pressure, and body temperature in static analysis data and real-time physical therapy data, including mean and variance values; time domain features can reflect the intensity and degree of change of data;

[0056] C2. Convert the time series data of pulse, blood pressure and body temperature into frequency domain data through Fourier transform, and extract the frequency domain features of the frequency domain data, including power spectrum density and spectrum diagram; the frequency domain features can reflect the frequency distribution of the data;

[0057] C3. Obtain the nonlinear characteristics of pulse, blood pressure, and body temperature time series data, including sample entropy and Lyapunov coefficient; nonlinear characteristics can reflect the complexity and dynamics of the signal;

[0058] By acquiring the time domain, frequency domain and nonlinear features of pulse, blood pressure and body temperature time series data, the number of features of smaller data can be increased and the potential value of physical therapy related data can be explored.

[0059] S3. The therapy object data is sent to the static data analysis model for analysis to obtain static data analysis results; the static data analysis results and the therapy object data are sent to the real-time therapy data analysis model for analysis to obtain real-time therapy data analysis results;

[0060] The static data analysis model includes a data processing layer and a static data analysis layer;

[0061] The data processing layer is used to pre-process the therapy object data to obtain static data to be analyzed;

[0062] The static data analysis layer is used to analyze the static data to be analyzed and output the static data analysis results;

[0063] The static data to be analyzed include basic physical data and characteristic data of static analysis data. Through these data, the model is used to analyze the user's health status, including the type of disease and severity of the disease;

[0064] Build a static data analysis model. The specific steps are:

[0065] A1. Obtain a number of physical therapy object data, preprocess the physical therapy object data and mark the static analysis result labels, form all the physical therapy object data that have been preprocessed and marked with static analysis result labels into a static analysis data set, and divide the static analysis data set into a static analysis training set and a static analysis test set;

[0066] A2. Send the static analysis training set to the initial static data analysis model based on the recurrent neural network for training, and use the static analysis result label as the target to train the optimized static data analysis model;

[0067] A3. Send the static analysis test set to the optimized static data analysis model for accuracy evaluation, and use the static analysis result label as the target to obtain the test accuracy; if the test accuracy is within the error standard range, the optimized static data analysis model is used as the static data analysis model; if the test accuracy exceeds the error standard range, adjust the hyperparameters of the optimized static data analysis model and repeat steps A2-A3;

[0068] A recurrent neural network is a neural network structure that can process sequence data. It is characterized by sharing weights between time steps and remembering previous information. The recurrent neural network can effectively obtain the potential connections between data and improve the recognition effect of the model. The initial static data analysis model contains 1 input layer, 5 hidden layers and 1 output layer. The Adam optimizer is used to automatically adjust the learning rate during training, and the ReLU activation function is used to avoid the problem of gradient disappearance in the recurrent neural network.

[0069] The real-time physical therapy data analysis model includes a data processing layer and a real-time data analysis layer;

[0070] The data processing layer is used to reconstruct the therapy object data according to the static data analysis results to obtain the reconstructed therapy object data, and preprocess the data to obtain the real-time therapy data to be analyzed;

[0071] The specific method of reconstruction is: replacing the features corresponding to the static analysis data with the static analysis results, and adding weights to the feature data corresponding to the real-time physical therapy data according to the static analysis results.

[0072] The real-time physical therapy data to be analyzed includes static analysis results, characteristic data of weighted real-time physical therapy data, physical therapy plans and basic body data;

[0073] The real-time data analysis layer is used to analyze the real-time physical therapy data to be analyzed and output the real-time physical therapy data analysis results;

[0074] Build a real-time physical therapy data analysis model. The specific steps are as follows:

[0075] B1. Obtain some physiotherapy object data and static data analysis results, reconstruct the physiotherapy object data according to the static data analysis results to obtain reconstructed physiotherapy object data, preprocess the reconstructed physiotherapy object data and mark the real-time physiotherapy data analysis result labels, form all the reconstructed physiotherapy object data that have been preprocessed and marked with real-time physiotherapy data analysis result labels into a real-time physiotherapy analysis data set, and divide the real-time physiotherapy analysis data set into a real-time physiotherapy analysis training set and a real-time physiotherapy analysis test set;

[0076] B2. Send the real-time physical therapy analysis training set to the initial real-time physical therapy data analysis model based on the long short-term memory network for training, and take the real-time physical therapy data analysis result label as the target to train and obtain the optimized real-time physical therapy data analysis model;

[0077] B3. Send the real-time physical therapy analysis test set to the optimized real-time physical therapy data analysis model for accuracy evaluation, and use the real-time physical therapy data analysis result label as the target to obtain the test accuracy; if the test accuracy is within the error standard range, the optimized real-time physical therapy data analysis model is used as the real-time physical therapy data analysis model; if the test accuracy exceeds the error standard range, adjust the hyperparameters of the optimized real-time physical therapy data analysis model and repeat steps B2-B3;

[0078] The use of long short-term memory networks improves the model's ability to handle long-term dependency problems. The physical therapy process data is a changing data, and the long short-term memory network can well capture the relationship between time series data. Compared with the common recurrent neural network, the initial real-time physical therapy data analysis model based on the long short-term memory network adds a forgetting mechanism, which automatically manages the data life cycle through the forgetting mechanism, ensuring that important information will not be permanently lost due to gradient disappearance or explosion problems.

[0079] S4. Matching a standard physiotherapy plan in the physiotherapy plan library according to the static data analysis results and the real-time physiotherapy data analysis results, generating physiotherapy suggestions, and outputting the standard physiotherapy plan and physiotherapy suggestions;

[0080] The matching method is: match the disease labels in the physiotherapy program library through the static data analysis results to obtain the disease matching results, match the physiotherapy program in the disease matching results through the real-time physiotherapy data analysis results, and output the matching standard physiotherapy program;

[0081] The therapy plan library stores the disease labels and therapy method labels corresponding to the therapy plans. The disease labels are matched by the disease types and degrees of the static data analysis results.

[0082] Therapeutic suggestions are stored in standard therapy plans and output after being intelligently adjusted accordingly based on the user’s basic body data;

[0083] S5. Collect relevant data of the user according to the set monitoring time, obtain static data analysis results and real-time physical therapy data analysis results of the relevant data of the user collected at that time, and use them as monitoring feedback data, adjust the physical therapy plan according to the monitoring feedback data, and generate feedback physical therapy suggestions;

[0084] The physiotherapy plan is adjusted according to the monitoring feedback data, and feedback physiotherapy suggestions are generated. The specific steps of the method are:

[0085] D1. The feedback data is matched again in the physiotherapy program library to obtain the feedback disease matching results and feedback physiotherapy programs;

[0086] D2. Adjust the physical therapy plan through the feedback physical therapy plan, and generate feedback physical therapy suggestions based on the feedback disease matching results and the feedback physical therapy plan.

[0087] S6. Perform a physiotherapy effect evaluation based on the static data analysis results of the relevant data of the user collected each time, and generate a physiotherapy effect evaluation table; the physiotherapy effect evaluation is obtained by comparing different static data analysis results, and the physiotherapy effect evaluation table is used to record all physiotherapy plans and physiotherapy effect evaluations of the user.

[0088] S7. Visualize the basic body data, physical therapy plan, physical therapy suggestions and physical therapy effect evaluation form, and output the visualization results.

[0089] Example 2, a physical therapy data intelligent analysis system, see Figure 1 As shown, including:

[0090] The data collection module is used to collect relevant data of the user, including gender, age, height, weight and disease type, as basic body data; time series data of pulse, blood pressure and body temperature in a static state, as static analysis data; time series data of pulse, blood pressure and body temperature during physical therapy and physical therapy methods, as real-time physical therapy data, pre-process the relevant data of the user, extract features from the pre-processed static analysis data and real-time physical therapy data, and organize the extracted features and basic body data into physical therapy object data;

[0091] The intelligent analysis module is used to send the therapy object data into the static data analysis model for analysis to obtain the static data analysis results; send the static data analysis results and the therapy object data into the real-time therapy data analysis model for analysis to obtain the real-time therapy data analysis results;

[0092] A physiotherapy plan generation module is used to match standard physiotherapy plans in the physiotherapy plan library according to the static data analysis results and the real-time physiotherapy data analysis results, generate physiotherapy suggestions, and output the standard physiotherapy plans and physiotherapy suggestions;

[0093] The monitoring feedback module is used to collect relevant data of the user according to the set monitoring time, obtain the static data analysis results and real-time physical therapy data analysis results of the relevant data of the user collected at this time, and use them as monitoring feedback data, adjust the physical therapy plan according to the monitoring feedback data, and generate feedback physical therapy suggestions;

[0094] The effect evaluation module is used to evaluate the physical therapy effect according to the static data analysis results of the relevant data of the user collected each time, and generate a physical therapy effect evaluation table;

[0095] The visualization interface module is used to visualize basic body data, physiotherapy plans, physiotherapy suggestions and physiotherapy effect evaluation tables, and output visualization results.

[0096] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all such improvements and changes should fall within the scope of protection of the appended claims of the present invention. Parts not described in detail in this specification belong to the prior art known to those skilled in the art.

Claims

1. A method for intelligent analysis of physical therapy data, characterized in that: include: S1. Collect relevant data of the user, including gender, age, height, weight and disease type, as basic physical data; time series data of pulse, blood pressure and body temperature in a static state, as static analysis data; time series data of pulse, blood pressure, body temperature and physical therapy methods during physical therapy, as real-time physical therapy data; S2. Preprocess the user's relevant data, extract features from the preprocessed static analysis data and real-time physical therapy data, and organize the extracted features and basic body data into physical therapy object data; S3. The therapy object data is sent to the static data analysis model for analysis to obtain static data analysis results; the reconstructed therapy object data is sent to the real-time therapy data analysis model for analysis to obtain real-time therapy data analysis results; S4. Matching a standard physiotherapy plan in the physiotherapy plan library according to the static data analysis results and the real-time physiotherapy data analysis results, generating physiotherapy suggestions, and outputting the standard physiotherapy plan and physiotherapy suggestions; S5. Collect relevant data of the user according to the set monitoring time, obtain static data analysis results and real-time physical therapy data analysis results of the relevant data of the user collected at that time, and use them as monitoring feedback data, adjust the physical therapy plan according to the monitoring feedback data, and generate feedback physical therapy suggestions; S6. Perform a physiotherapy effect evaluation based on the static data analysis results of the user's relevant data collected each time, and generate a physiotherapy effect evaluation table; S7. Visualize the basic body data, physical therapy plan, physical therapy suggestions and physical therapy effect evaluation table, and output the visualization results; The real-time physical therapy data analysis model includes a data processing layer and a real-time data analysis layer; The data processing layer is used to reconstruct the therapy object data according to the static data analysis results to obtain the reconstructed therapy object data, and preprocess the data to obtain the real-time therapy data to be analyzed; The real-time data analysis layer is used to analyze the real-time physical therapy data to be analyzed and output the real-time physical therapy data analysis results; Build a real-time physical therapy data analysis model. The specific steps are as follows: B1. Obtain some physiotherapy object data and static data analysis results, reconstruct the physiotherapy object data according to the static data analysis results to obtain reconstructed physiotherapy object data, preprocess the reconstructed physiotherapy object data and mark the real-time physiotherapy data analysis result labels, form all the reconstructed physiotherapy object data that have been preprocessed and marked with real-time physiotherapy data analysis result labels into a real-time physiotherapy analysis data set, and divide the real-time physiotherapy analysis data set into a real-time physiotherapy analysis training set and a real-time physiotherapy analysis test set; B2. Send the real-time physical therapy analysis training set to the initial real-time physical therapy data analysis model based on the long short-term memory network for training, and take the real-time physical therapy data analysis result label as the target to train and obtain the optimized real-time physical therapy data analysis model; B3. Send the real-time physical therapy analysis test set to the optimized real-time physical therapy data analysis model for accuracy evaluation, using the real-time physical therapy data analysis result label as the target to obtain the test accuracy; If the test accuracy is within the error standard range, the optimized real-time physical therapy data analysis model is used as the real-time physical therapy data analysis model. If the test accuracy exceeds the error standard range, the hyperparameters of the optimized real-time physical therapy data analysis model are adjusted and steps B2-B3 are repeated. Reconstructing the therapy object data according to the static data analysis results to obtain reconstructed therapy object data, wherein the specific method of reconstruction is: replacing the features corresponding to the static analysis data with the static analysis results, and adding weights to the feature data corresponding to the real-time therapy data according to the static analysis results; Feature extraction is performed on the preprocessed static analysis data and real-time physical therapy data. The specific steps of feature extraction are as follows: C1. Extract time domain features from the time series data of pulse, blood pressure, and body temperature in static analysis data and real-time physical therapy data, including mean and variance values; C2. Convert the time series data of pulse, blood pressure and body temperature into frequency domain data through Fourier transform, and extract the frequency domain features of the frequency domain data, including power spectrum density and spectrum diagram; C3. Obtain the nonlinear characteristics of time series data of pulse, blood pressure, and body temperature, including sample entropy and Lyapunov coefficient.

2. A method for intelligent analysis of physical therapy data according to claim 1, characterized in that: The static data analysis model includes a data processing layer and a static data analysis layer; The data processing layer is used to pre-process the therapy object data to obtain static data to be analyzed; The static data analysis layer is used to analyze the static data to be analyzed and output the static data analysis results; Build a static data analysis model. The specific steps are: A1. Obtain a number of physical therapy object data, preprocess the physical therapy object data and mark the static analysis result labels, form all the physical therapy object data that have been preprocessed and marked with static analysis result labels into a static analysis data set, and divide the static analysis data set into a static analysis training set and a static analysis test set; A2. Send the static analysis training set to the initial static data analysis model based on the recurrent neural network for training, and use the static analysis result label as the target to train the optimized static data analysis model; A3. Send the static analysis test set to the optimized static data analysis model for accuracy evaluation, and use the static analysis result label as the target to obtain the test accuracy; If the test accuracy is within the error standard range, the optimized static data analysis model is used as the static data analysis model. If the test accuracy exceeds the error standard range, the hyperparameters of the optimized static data analysis model are adjusted and steps A2-A3 are repeated.

3. A method for intelligent analysis of physical therapy data according to claim 1, characterized in that: According to the static data analysis results and the real-time physical therapy data analysis results, the standard physical therapy plan is matched in the physical therapy plan library. The matching method is: match the disease label in the physical therapy plan library through the static data analysis results to obtain the disease matching results, match the physical therapy plan in the disease matching results through the real-time physical therapy data analysis results, and output the matched standard physical therapy plan.

4. A method for intelligent analysis of physical therapy data according to claim 1, characterized in that: The physiotherapy plan is adjusted according to the monitoring feedback data, and feedback physiotherapy suggestions are generated. The specific steps of the method are: D1. The feedback data is matched again in the physiotherapy program library to obtain the feedback disease matching results and feedback physiotherapy programs; D2. Adjust the physical therapy plan through the feedback physical therapy plan, and generate feedback physical therapy suggestions based on the feedback disease matching results and the feedback physical therapy plan.

5. The method for intelligent analysis of physical therapy data according to claim 1, characterized in that: The physiotherapy effect evaluation is performed based on the static data analysis results of the relevant data of the user collected each time, and a physiotherapy effect evaluation table is generated; the physiotherapy effect evaluation is obtained by comparing different static data analysis results, and the physiotherapy effect evaluation table is used to record all physiotherapy plans and physiotherapy effect evaluations of the user.

6. A physical therapy data intelligent analysis system, characterized in that: The system applies a method for intelligent analysis of physical therapy data according to any one of claims 1 to 5, including: The data collection module is used to collect relevant data of the user, including gender, age, height, weight and disease type, as basic body data; time series data of pulse, blood pressure and body temperature in a static state, as static analysis data; time series data of pulse, blood pressure and body temperature during physical therapy and physical therapy methods, as real-time physical therapy data, pre-process the relevant data of the user, extract features from the pre-processed static analysis data and real-time physical therapy data, and organize the extracted features and basic body data into physical therapy object data; The intelligent analysis module is used to send the therapy object data into the static data analysis model for analysis to obtain the static data analysis results; send the static data analysis results and the therapy object data into the real-time therapy data analysis model for analysis to obtain the real-time therapy data analysis results; A physiotherapy plan generation module is used to match standard physiotherapy plans in the physiotherapy plan library according to the static data analysis results and the real-time physiotherapy data analysis results, generate physiotherapy suggestions, and output the standard physiotherapy plans and physiotherapy suggestions; The monitoring feedback module is used to collect relevant data of the user according to the set monitoring time, obtain the static data analysis results and real-time physical therapy data analysis results of the relevant data of the user collected at this time, and use them as monitoring feedback data, adjust the physical therapy plan according to the monitoring feedback data, and generate feedback physical therapy suggestions; The effect evaluation module is used to evaluate the physical therapy effect according to the static data analysis results of the relevant data of the user collected each time, and generate a physical therapy effect evaluation table; The visualization interface module is used to visualize basic body data, physiotherapy plans, physiotherapy suggestions and physiotherapy effect evaluation tables, and output visualization results.

Images