Time sequence fusion and health state prediction method and system for multi-dimensional physiological data

Abstract

The invention provides a time sequence fusion and health state prediction method and system for multi-dimensional physiological data, and relates to the technical field of electrical digital data processing.The method includes the steps that high-precision sensing equipment and a parameter calibration model are adopted to calibrate data, time synchronization and error compensation are achieved through a weighted fusion algorithm, and the accuracy of physiological parameters is improved; meanwhile, a fatigue index and a pressure index are corrected in real time by using a closed-loop feedback mechanism, and a comprehensive state index is generated based on an evaluation result for further analysis and prediction; the change trend of the comprehensive state index is predicted through a machine learning technology, and a targeted dynamic rehabilitation scheme including training intensity optimization, diet adjustment, rest cycle planning and the like is provided in combination with physiological data of the user; and a timing feedback mechanism is established, so that the user can obtain health state change and adjustment suggestions in real time, thereby improving scientificity and timeliness of a rehabilitation scheme, remarkably optimizing the rehabilitation effect and efficiency, and realizing intelligent health management at the same time.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic digital data processing, and in particular to a method and system for time series fusion and health status prediction of multi-dimensional physiological data. Background Art

[0002] Exercise rehabilitation status estimation methods originated in sports medicine, aiming to assess a patient's recovery progress. With the development of exercise physiology and biomechanics, scientific assessment tools, such as functional assessment criteria and biomarkers, have gradually emerged, promoting personalized and precise rehabilitation treatment.

[0003] In the prior art, the publication number is CN106570346A, and the name is Physiological Condition Assessment Factor Determination Method and Physiological Condition Assessment Factor Determination System. The physiological condition assessment factor determination method includes the following steps: obtaining multiple first risk factors for a certain physiological condition based on knowledge; obtaining multiple second risk molecules related to the same physiological condition based on clinical data; performing a logistic regression model analysis on the second risk molecules; calculating the correlation coefficient between the first risk molecule and the second risk molecule after the logistic regression model analysis to determine the correlation between the first risk molecule and the second risk molecule; screening out assessment factors for a certain physiological condition based on the correlation between the first risk molecule and the second risk molecule to determine valuable assessment factors for a certain physiological condition. It can effectively improve the quality of risk factors, reduce the modeling dimensions of risk prediction models, provide effective risk factor assessments, and provide an effective basis for risk prediction of physiological conditions.

[0004] The above technical solutions often have the following technical shortcomings when applied to physiological status assessment methods for postoperative rehabilitation: The system lacks real-time dynamic collection and integration of multi-dimensional physiological data under multiple motion states, and focuses on the screening and assessment of static risk factors. It is unable to achieve automatic calibration, time synchronization, and error compensation for key physiological indicators such as fatigue and stress under different motion states. This leads to insufficient capture of dynamic changes in the user's health status and difficulty in accurately reflecting the actual physiological condition. At the same time, the above technical solutions lack the closed-loop feedback correction and comprehensive evaluation and prediction functions of health status based on time series data; due to the failure to achieve real-time closed-loop adjustment of fatigue index and stress index, and the failure to build an analysis model of the comprehensive status index and its changing trend, it is impossible to provide users with dynamic health status prediction and personalized rehabilitation plan guidance, which limits the timeliness and personalization level of health management.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the Invention

[0006] The purpose of the present invention is to provide a method and system for time series fusion and health status prediction of multi-dimensional physiological data to solve the problems raised in the above background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions: A method for time series fusion and health status prediction of multi-dimensional physiological data, comprising the following steps: Step 1: Real-time collection of multi-dimensional physiological data of the user at different exercise states during the current exercise and rehabilitation phase, and analysis of the multi-dimensional physiological data at different exercise states to construct a parameter calibration model; the multi-dimensional physiological data includes fatigue-related indicators and stress-related indicators; The parameter calibration model is used to automatically calibrate multidimensional physiological data of different motion states; Step 2: Obtain multidimensional physiological data after automatic calibration, use a weighted fusion algorithm to synchronize time and compensate for errors in the multidimensional physiological data of different motion states, and generate physiological parameter sequences of different motion states; Step 3: Based on the physiological parameter sequences of different exercise states, the fatigue index and stress index under different exercise states are calculated respectively, and closed-loop correction is performed; a threshold is preset, and the corrected fatigue index and stress index are compared and evaluated; Step 4: Based on the evaluation contents of fatigue index and stress index, further generate comprehensive state index under different exercise states; At the same time, the comprehensive state indexes under adjacent motion states were analyzed to generate a difference index; Step 5: Based on the difference index, predict the changing trend of the comprehensive status index within a preset time in the future; set a feedback cycle, and regularly provide advance feedback on the changing trend of the comprehensive status index to the user through wearable devices, and at the same time issue dynamic rehabilitation plan adjustment suggestions.

[0008] Furthermore, step one specifically includes: Wearable devices are configured to collect multi-dimensional physiological data in real time, including heart rate (HR), respiratory rate (BR), blood oxygen saturation (Xyb), heart rate variability (HRV), galvanic skin response (GSR), and cortisol level (Cor). These data are then transmitted to the wearable device's main control unit via BLE communication for unified processing. Timestamps are added to ensure that the multi-dimensional physiological data has temporal correlation. Based on the motion analysis results of the three-axis accelerometer in the wearable device, the different consecutive motion states in the current exercise and rehabilitation stage are sequentially marked to form a marking sequence {1, 2, ..., i, ..., n}, where i represents the sequential mark of the motion state at the current moment, and n represents the total number of motion states finally identified in the entire exercise and rehabilitation stage; According to the tag sequence, the collected multidimensional physiological data are associated with the corresponding motion state to form a multidimensional data set.

[0009] Furthermore, step one specifically includes: During the user's exercise, the main control MCU reads the original multi-dimensional physiological data currently marked as i in real time and outputs the calibration value through the trained vector regression model; If a change in the marker sequence is detected, that is, the motion state switches from i to i+1, the online fine-tuning mechanism is triggered, and incremental learning is performed using real-time physiological data and the marker sequence to update the model parameters; The calibrated data is transmitted to the application in real time via wireless communication.

[0010] Furthermore, step 2 specifically includes: A weighted fusion algorithm is used to perform error compensation and time synchronization on the automatically calibrated multi-dimensional physiological data. Data with inconsistent timestamps are synchronized using linear interpolation, including calculation of interpolated heart rate data. Use interpolation methods to synchronize data with different timestamps to compensate for signal loss or delays during the acquisition process. Select linear interpolation to achieve smooth transition of time data and ensure the temporal continuity of the fused data. Error compensation is performed on the collected data. Error calculation is based on the statistical difference of historical data and dynamic error correction is performed using the Kalman filter. The Kalman filter dynamically adjusts the error estimate based on the difference between the real-time observation value and the predicted value. The generated physiological parameter sequence is processed by time synchronization and error compensation to form a continuous physiological parameter sequence {P1, P2, ..., P i ,…,P n}, each value represents the fusion physiological parameter value under a motion state, where P i represents the physiological parameter value of the i-th motion state, P n The physiological parameter value representing the nth motion state; Furthermore, step three specifically includes: Based on the heart rate, respiratory rate and blood oxygen saturation in the physiological parameter sequence, after standardization, {HR i ,BR i ,Xyb i} respectively represent the heart rate, respiratory rate and blood oxygen saturation in the i-th exercise state. The fatigue index is calculated using the following formula:

[0011] in, is the fatigue index under the i-th exercise state, k1, k2 and k3 correspond to the weight coefficients of heart rate, respiratory rate and blood oxygen saturation respectively, and k1+k2+k3=1; Based on the heart rate variability, skin galvanic response and cortisol level in the physiological parameter sequence, standardization is performed and {HRV i ,GSR i ,Cor i} respectively represent the heart rate variability, skin galvanic response and cortisol level in the i-th exercise state. The stress index is calculated using the following formula:

[0012] in, is the stress index under the i-th exercise state, k4, k5 and k6 are the weight coefficients of the corresponding heart rate variability, skin electrode response and cortisol level, and k4+k5+k6=1.

[0013] Furthermore, step three specifically includes: Perform real-time closed-loop feedback correction on fatigue index and stress index, extract two consecutive timestamps t from the data stream 1 and the fatigue index corresponding to t, namely Apl(t 1) and Apl(t), and the pressure index, Ayl(t 1) and Ayl(t), and obtain fatigue deviation ΔApl and pressure deviation ΔAyl respectively through difference calculation; The specific content of the closed-loop feedback adjustment includes comprehensive evaluation and visual output of the corrected fatigue index and stress index; the evaluation logic is as follows: Set the fatigue index and stress index thresholds, namely fatigue threshold Tpl and stress threshold Tyl respectively; Normal range: If both the fatigue index and stress index are below their respective thresholds, it is considered normal and the user can continue with the current activity. Mild Abnormality: If either the fatigue index or the stress index exceeds the corresponding threshold, a mild abnormality warning will be issued and suggestions for adjusting activities will be made; Severe abnormality: If both the fatigue index and stress index exceed the corresponding thresholds, it is marked as a severe abnormality. At this time, the user is advised to adjust the exercise intensity in time or require necessary health checks.

[0014] Furthermore, step four specifically includes: The fatigue index and stress index are integrated to generate a comprehensive status index. The specific calculation formula is as follows:

[0015] in, represents the comprehensive state index under the current motion state i, k7 and k8 are the weight parameters of the fusion fatigue index and pressure index respectively, and k7+k8=1; After generating the comprehensive state index, the difference index under adjacent motion states is analyzed by comparing the difference in the comprehensive state index between adjacent motion states i and i+1, and generating the difference index , the specific calculation formula is as follows:

[0016] Difference Index It reflects the change in stress and fatigue from the current state to the next state. Furthermore, step five specifically includes: The comprehensive state index of the above motion state is established by machine learning method and difference index As input, a state prediction model is used to predict the trend of the comprehensive state index in the future; Set the timing of data synchronization and feedback cycles based on the user's specific needs and activity types; At the end of each cycle, the prediction result of the comprehensive state index is transmitted to the data transmission module on the user's wearable device; Among them, the prediction techniques of machine learning methods include linear regression and time series analysis.

[0017] Furthermore, step five specifically includes: Automatically generate personalized rehabilitation or training adjustment suggestions based on the predicted trend of the comprehensive status index. By analyzing each user's historical health data, current health status, and predicted future health trends, targeted suggestions are generated based on this analysis, including adjusting training intensity, optimizing diet plans, and adjusting rest periods. In addition, a user feedback mechanism is established to allow users to provide feedback on the suggestions received, which can be used to optimize the prediction model and adjust future health management strategies.

[0018] A multi-dimensional physiological data time series fusion and health status prediction system, comprising: The physiological data acquisition module collects multi-dimensional physiological data of the user in different exercise states in real time during the current exercise and rehabilitation stage, and analyzes the multi-dimensional physiological data of different exercise states to build a parameter calibration model; the multi-dimensional physiological data includes fatigue-related indicators and stress-related indicators; The parameter calibration model is used to automatically calibrate multidimensional physiological data of different motion states; The data fusion and calibration module acquires multi-dimensional physiological data after automatic calibration, uses a weighted fusion algorithm to synchronize time and compensate for errors in the multi-dimensional physiological data of different motion states, and generates physiological parameter sequences of different motion states; The index calculation and correction module calculates the fatigue index and stress index under different exercise states based on the physiological parameter sequences of different exercise states, and performs closed-loop correction. It presets thresholds and compares and evaluates the corrected fatigue index and stress index. The state index analysis module further generates comprehensive state indices under different exercise states based on the evaluation content of fatigue index and stress index; At the same time, the comprehensive state indexes under adjacent motion states were analyzed to generate a difference index; The trend prediction and feedback module predicts the changing trends of the comprehensive status index for multiple times within a preset time in the future based on the difference index; and sets a feedback cycle to regularly provide advance feedback on the changing trends of the comprehensive status index to users through wearable devices, while also issuing dynamic rehabilitation plan adjustment suggestions.

[0019] Compared with the prior art, the present invention has the following beneficial effects: The present invention configures wearable devices to collect multidimensional physiological data in real time, and uses a three-axis accelerometer to sequentially mark different motion states to form a mark sequence, thereby realizing the dynamic collection and association of multidimensional physiological data under multiple motion states. It uses a trained vector regression model to automatically calibrate the original multidimensional physiological data, combines an online fine-tuning mechanism to dynamically update the model parameters, and uses a weighted fusion algorithm, linear interpolation, and Kalman filter to synchronize the data and compensate for errors, generating a continuous and accurate physiological parameter sequence. This effectively solves the problems of insufficient dynamic change capture and imperfect data fusion in the existing technology. The present invention calculates fatigue index and stress index, adjusts fatigue index and stress index in real time through closed-loop feedback correction mechanism, performs comprehensive evaluation based on preset thresholds and generates comprehensive state index under fusion weights k7 and k8, further calculates difference index between adjacent motion states, establishes state prediction model with comprehensive state index and difference index as input based on machine learning method, predicts future trend of comprehensive state index change and feeds back to users through wearable devices, automatically generates personalized rehabilitation and training adjustment suggestions, realizes dynamic closed-loop management and personalized guidance of health status, and significantly improves the timeliness and personalization level of health management. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 Schematic diagram of the overall method flow of the present invention; Figure 2 It is a schematic diagram of the overall system framework of the present invention. DETAILED DESCRIPTION

[0021] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to specific embodiments.

[0022] It should be noted that, unless otherwise defined, the technical or scientific terms used in the present invention should have the usual meanings understood by people with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative position relationships. When the absolute position of the object being described changes, the relative position relationship will also change accordingly.

[0023] Example 1: See also Figure 1 The present invention provides a technical solution: a method for time series fusion and health status prediction of multi-dimensional physiological data, which specifically includes the following steps: Step 1: Real-time collection of multi-dimensional physiological data of the user at different exercise states during the current exercise and rehabilitation phase, and analysis of the multi-dimensional physiological data at different exercise states to construct a parameter calibration model; the multi-dimensional physiological data includes fatigue-related indicators and stress-related indicators; The parameter calibration model is used to automatically calibrate multidimensional physiological data of different motion states; Step 2: Obtain multidimensional physiological data after automatic calibration, use a weighted fusion algorithm to synchronize time and compensate for errors in the multidimensional physiological data of different motion states, and generate physiological parameter sequences of different motion states; Step 3: Based on the physiological parameter sequences of different exercise states, the fatigue index and stress index under different exercise states are calculated respectively, and closed-loop correction is performed; a threshold is preset, and the corrected fatigue index and stress index are compared and evaluated; Step 4: Based on the evaluation contents of fatigue index and stress index, further generate comprehensive state index under different exercise states; At the same time, the comprehensive state indexes under adjacent motion states were analyzed to generate a difference index; Step 5: Based on the difference index, predict the changing trend of the comprehensive status index within a preset time in the future; set a feedback cycle, and regularly provide advance feedback on the changing trend of the comprehensive status index to the user through wearable devices, and at the same time issue dynamic rehabilitation plan adjustment suggestions.

[0024] Step 1 specifically includes: The wearable device is configured to collect multi-dimensional physiological data including heart rate (HR), respiratory rate (BR), blood oxygen saturation (Xyb), heart rate variability (HRV), galvanic skin response (GSR), and cortisol level (Cor) in real time, and transmit the data in real time to the main control unit of the wearable device via BLE communication for unified processing. At the same time, timestamps are added to make the multi-dimensional physiological data time-correlated. In this embodiment, "BLE" represents "Bluetooth-Low-Energy" low-power Bluetooth technology. Based on the motion analysis results of the three-axis accelerometer in the wearable device, the different consecutive motion states in the current exercise and rehabilitation stage are sequentially marked to form a marking sequence {1, 2, ..., i, ..., n}, where i represents the sequential mark of the motion state at the current moment, and n represents the total number of motion states finally identified in the entire exercise and rehabilitation stage; Furthermore, the specific operations for collecting multi-dimensional physiological data include: Heart rate and blood oxygen data are collected using a photoplethysmography (PPG) sensor. A three-axis accelerometer is used to measure motion status information, such as stillness, walking, and running. A chest sensor is used to monitor respiratory rate. Muscle activity is measured using electrode-based sensors. A continuous sequence of markers is associated with the collected data to indicate the user's current motion phase. Furthermore, it is assumed that during the user's exercise rehabilitation phase, the exercise state recognition algorithm is used to classify and sequentially mark the exercise state: The user starts from "Standstill", with state marked as 1; Then the user enters the "walking" state, marked as 2; Then enter the "running" state, marked as 3; Finally, enter the "rest" state, marked as 4; In this case, i represents the current label of the motion state, that is, at a certain moment, the user is in the motion state labeled 2, which is “walking”; n represents the total number of all different motion states identified. In this example, n=4, including the motion states of stillness, walking, running, and resting. The structure of the marker sequence can help the model accurately track the changes in the user's motion state, and in the subsequent analysis and calibration process, associate the physiological data of each state with its motion stage, thus forming a complete data recording and analysis process.

[0025] According to the tag sequence, the collected multidimensional physiological data are associated with the corresponding motion state to form a multidimensional data set.

[0026] Use physiological data parameters to conduct preliminary analysis of exercise status and user physiological performance: A dynamic threshold algorithm is used to assess changes in heart rate (HR) and blood oxygen saturation (Xyb) during specific exercise states to determine fatigue or stress. Muscle activity levels are analyzed based on electromyography (EMG) data, and exercise intensity is determined using a three-axis accelerometer (ACC). Utilize state recognition algorithms, including decision trees, to classify motion states and continuously update tag sequences, thereby evaluating the user's continuous motion state in real time. Combining the marker sequence {1, 2, ..., i, ..., n} and the corresponding physiological data set, a calibration model of motion state and physiological parameters is established through regression analysis or machine learning algorithm: For the calibration of physiological indicators, multiple linear regression was used:

[0027] Where Y represents the quantitative index of exercise status, i.e. fatigue level or exercise performance score; HR stands for heart rate variable, which is the heart rate value measured under specific exercise conditions; Xyb represents the blood oxygen saturation variable, that is, the blood oxygen saturation measured under a specific exercise state; α represents the intercept in the regression model, which represents the expected value of the target variable when all independent variables are zero; β1 and β2 represent regression coefficients, which respectively describe the influence of heart rate and blood oxygen saturation on the quantitative indicators of exercise state; is the error term, which represents the deviation between the observed value and the regression line, reflecting the influence of other unobserved factors except heart rate and blood oxygen saturation; The operating logic of this formula is to collect corresponding heart rate and blood oxygen saturation data, as well as corresponding exercise status scores or labels, under different exercise states; use statistical analysis software or programming environments, such as R language or Python, to apply the collected data to the above regression model equation, and calculate the regression parameters of the optimal solution through the least squares method; model verification: cross-validate the model to check its prediction accuracy and generalization ability; this step can help confirm whether the model is effective and its reliability; after the model is established, it can be used to predict the exercise status of the athlete under given heart rate and blood oxygen saturation, or provide personalized exercise recommendations for the athlete.

[0028] Step 1 specifically includes: During the user's exercise, the microcontroller-unit (MCU) reads the raw multi-dimensional physiological data currently marked as i in real time and outputs calibration values ​​through the trained vector regression model; If a change in the marker sequence is detected, that is, the motion state switches from i to i+1, the online fine-tuning mechanism is triggered, and incremental learning is performed using real-time physiological data and the marker sequence to update the model parameters; The calibrated data is transmitted to the application in real time via wireless communication.

[0029] Step 2 specifically includes: A weighted fusion algorithm is used to perform error compensation and time synchronization on the automatically calibrated multi-dimensional physiological data. Data with inconsistent timestamps are synchronized using linear interpolation, including calculation of interpolated heart rate data. Use interpolation methods to synchronize data with different timestamps to compensate for signal loss or delays during the acquisition process. Select linear interpolation to achieve smooth transition of time data and ensure the temporal continuity of the fused data. Error compensation is performed on the collected data. Error calculation is based on the statistical difference of historical data and dynamic error correction is performed using the Kalman filter. The Kalman filter dynamically adjusts the error estimate based on the difference between the real-time observation value and the predicted value. The generated physiological parameter sequence is processed by time synchronization and error compensation to form a continuous physiological parameter sequence {P1, P2, ..., P i ,…,P n}, each value represents the fusion physiological parameter value under a motion state, where P i represents the physiological parameter value of the i-th motion state, P n The physiological parameter value representing the nth motion state; Application of weighted fusion algorithm: To achieve the fusion of different data sources, a weighted fusion model is defined, which involves adjusting the weight coefficients of multiple data sources; the weight setting is optimized based on the data source acquisition quality and its correlation with the target motion state; Based on the formula:

[0030] Where: P i is the physiological parameter value of the i-th motion state; M3 is the number of data source types, namely heart rate, blood oxygen saturation and respiratory rate; w k is the weight coefficient corresponding to the kth data source; X k,i is the original multidimensional physiological data collected by the kth data source under the i-th motion state; In this formula, the weight coefficient w k and data source X k,i The specific number and type of data depend on the system settings and the actual collected data, rather than describing individual physiological indicators one by one, thus simplifying the form to adapt to the scalability and flexibility of multi-source data; In this embodiment, the specific process of the physiological parameter sequence undergoing time synchronization and error compensation is as follows: During the processing of physiological parameter sequences, linear interpolation is first used to correct inconsistent timestamps between different data sources, addressing signal loss or delays that may occur during sensor acquisition and ensuring temporal continuity of the data. The interpolation formula calculates the difference between timestamps to generate a smoothly interpolated data set. Next, a Kalman filter is used to dynamically compensate for the interpolated data, adjusting the error between the observed and predicted values ​​in real time to achieve dynamic correction of data accuracy. The Kalman filter updates the Kalman gain and combines the prediction and observation errors to generate corrected, accurate physiological parameter values. Finally, a weighted fusion algorithm is used to adjust the weight coefficients based on the data source acquisition quality and correlation with the movement state. The fused multi-source data forms a continuous physiological parameter sequence that is temporally consistent and fully error-compensated, ensuring high data accuracy and real-time performance, providing reliable support for subsequent rehabilitation assessments and program adjustments.

[0031] Step three specifically includes: Based on the heart rate, respiratory rate and blood oxygen saturation in the physiological parameter sequence, after standardization, {HR i ,BR i ,Xyb i} respectively represent the heart rate, respiratory rate and blood oxygen saturation in the i-th exercise state. The fatigue index is calculated using the following formula:

[0032] in, is the fatigue index under the i-th exercise state, k1, k2 and k3 correspond to the weight coefficients of heart rate, respiratory rate and blood oxygen saturation respectively, and k1+k2+k3=1; Furthermore, the normalized value range of the heart rate, respiratory rate and blood oxygen saturation of the i-th exercise state is (0,1); Based on the heart rate variability, skin galvanic response and cortisol level in the physiological parameter sequence, standardization is performed and {HRV i ,GSR i ,Cor i} respectively represent the heart rate variability, skin galvanic response and cortisol level in the i-th exercise state. The stress index is calculated using the following formula:

[0033] in, is the stress index under the i-th exercise state, k4, k5 and k6 are the weight coefficients of the corresponding heart rate variability, skin electrode response and cortisol level, and k4+k5+k6=1.

[0034] Step three specifically includes: Perform real-time closed-loop feedback correction on fatigue index and stress index, extract two consecutive timestamps t from the data stream 1 and the fatigue index corresponding to t, namely Apl(t 1) and Apl(t), and the pressure index, Ayl(t 1) and Ayl(t), and obtain fatigue deviation ΔApl and pressure deviation ΔAyl respectively through difference calculation; The specific calculation formula for the deviation is:

[0035] Based on the fatigue deviation ΔApl and the pressure deviation ΔAyl, real-time closed-loop feedback correction is performed. Based on the hierarchical analysis method used by the expert group, warning thresholds are set for the fatigue index and the pressure index, which are defined as fatigue threshold Tpl and pressure threshold Tyl respectively. The specific content of the closed-loop feedback adjustment includes comprehensive evaluation and visual output of the corrected fatigue index and stress index; the evaluation logic is as follows: Normal range: If both the fatigue index and stress index are below their respective thresholds, namely the fatigue threshold Tpl and the stress threshold Tyl, then the user is considered normal and can continue with their current activity. Mild Abnormality: If either the fatigue index or the stress index exceeds the corresponding threshold, a mild abnormality warning will be issued and suggestions for adjusting activities will be made; Severe abnormality: If both the fatigue index and stress index exceed the corresponding thresholds, it is marked as a severe abnormality. At this time, the user is advised to adjust the exercise intensity in time or require necessary health checks.

[0036] Step 4 specifically includes: The fatigue index and stress index are integrated to generate a comprehensive status index. The specific calculation formula is as follows:

[0037] in, represents the comprehensive state index under the current motion state i, k7 and k8 are the weight parameters of the fusion fatigue index and pressure index respectively, and k7+k8=1; After generating the comprehensive state index, the difference index under adjacent motion states is analyzed by comparing the difference in the comprehensive state index between adjacent motion states i and i+1, and generating the difference index , the specific calculation formula is as follows:

[0038] Difference Index This method reflects the magnitude of change in stress and fatigue from the current state to the next. Monitoring fatigue and stress in athletes or users across different exercise states is crucial in sports training and health management. This method provides a concise and effective way to integrate these indicators. By analyzing changes in adjacent states, it helps coaches and athletes better understand the dynamic changes in their physical fitness and emotions, thereby optimizing training and recovery strategies. This approach is superior to other complex models such as neural networks because it is not only computationally simple but also highly transparent, making it easy for athletes and coaches to understand and use.

[0039] Step 5 specifically includes: The comprehensive state index of the above motion state is established by machine learning method and difference index As input, a state prediction model is used to predict the trend of the comprehensive state index in the future; Set the timing of data synchronization and feedback cycles based on the user's specific needs and activity types; At the end of each cycle, the prediction result of the comprehensive state index is transmitted to the data transmission module on the user's wearable device; Among them, the prediction techniques of machine learning methods include linear regression and time series analysis.

[0040] Step 5 specifically includes: Automatically generate personalized rehabilitation or training adjustment suggestions based on the predicted trend of the comprehensive status index. By analyzing each user's historical health data, current health status, and predicted future health trends, targeted suggestions are generated based on this analysis, including adjusting training intensity, optimizing diet plans, and adjusting rest periods. In addition, a user feedback mechanism is established to allow users to provide feedback on the suggestions received, which can be used to optimize the prediction model and adjust future health management strategies.

[0041] In this embodiment, to meet the needs of real-time patient health status prediction and personalized rehabilitation recommendations in postoperative rehabilitation management, this technical solution uses machine learning to establish a status prediction model. Using a comprehensive status index and a difference index as input variables, this model predicts the changing trend of the patient's comprehensive status index over a period of time based on linear regression and time series analysis. The prediction model's calculation formula optimizes model parameters using a training dataset, thereby achieving accurate prediction of the patient's health status trend.

[0042] Based on the status prediction results, a data synchronization and feedback cycle is set according to the patient's specific activity type and recovery status. At the end of the cycle, the predicted comprehensive status index results are fed back to the terminal device worn by the patient through the data transmission module, ensuring real-time data transmission and facilitating the patient to view and track their health status. Subsequently, based on the changing trends of the comprehensive status index, combined with the patient's historical health data and current health status, personalized rehabilitation recommendations are generated for the patient. These recommendations include adjusting training intensity, optimizing diet plans and rest periods, etc., providing comprehensive and targeted support for the patient's postoperative recovery.

[0043] To enhance the accuracy of the prediction model and user experience, this solution also incorporates a user feedback mechanism. After receiving health recommendations, patients can provide feedback on their devices. This feedback is used to dynamically optimize the machine learning model and, in turn, refine future health management strategies for the patient. This closed-loop feedback loop further enhances the personalization of rehabilitation recommendations and the reliability of model predictions, forming a health management system that adjusts and continuously optimizes in real time.

[0044] The overall technical solution effectively meets the needs for accurate prediction and dynamic suggestions in postoperative rehabilitation management through status prediction and real-time adjustment mechanisms, while ensuring the feasibility and innovation of the model. Its results can significantly improve patients' rehabilitation experience and enhance rehabilitation quality.

[0045] Example 2: See also Figure 2 , a multi-dimensional physiological data time series fusion and health status prediction system, including: The physiological data acquisition module collects multi-dimensional physiological data of the user in different exercise states in real time during the current exercise and rehabilitation stage, and analyzes the multi-dimensional physiological data of different exercise states to build a parameter calibration model; the multi-dimensional physiological data includes fatigue-related indicators and stress-related indicators; The parameter calibration model is used to automatically calibrate multidimensional physiological data of different motion states; The data fusion and calibration module acquires multi-dimensional physiological data after automatic calibration, uses a weighted fusion algorithm to synchronize time and compensate for errors in the multi-dimensional physiological data of different motion states, and generates physiological parameter sequences of different motion states; The index calculation and correction module calculates the fatigue index and stress index under different exercise states based on the physiological parameter sequences of different exercise states, and performs closed-loop correction. It presets thresholds and compares and evaluates the corrected fatigue index and stress index. The state index analysis module further generates comprehensive state indices under different exercise states based on the evaluation content of fatigue index and stress index; At the same time, the comprehensive state indexes under adjacent motion states were analyzed to generate a difference index; The trend prediction and feedback module predicts the changing trends of the comprehensive status index for multiple times within a preset time in the future based on the difference index; and sets a feedback cycle to regularly provide advance feedback on the changing trends of the comprehensive status index to users through wearable devices, while also issuing dynamic rehabilitation plan adjustment suggestions.

[0046] Based on the wearable device-based exercise rehabilitation status estimation system in this embodiment, the following is a simulated operation example of the wearable device-based exercise rehabilitation status estimation system, demonstrating the device's use in exercise management for postoperative rehabilitation patients: Table 1 shows the physiological data of users in different exercise states:

[0047] The automatically calibrated physiological data were analyzed using a weighted fusion algorithm and time synchronization was achieved using linear interpolation. The weighted fusion algorithm is used to calculate the physiological parameter value P of the i-th motion state i : P i =0.3 0.6+0.2 0.5+0.5 0.8=0.67; Calculate the fatigue index under the i-th motion state , the specific calculation formula is:

[0048] Calculate the pressure index in the i-th motion state. The specific calculation formula is:

[0049] The calculation formula of the comprehensive state index under motion state i is:

[0050] Difference Index The calculation formula is: =0.415 0.37; Trend prediction and feedback: Use machine learning methods to predict the trend of changes in the comprehensive status index and provide user feedback; Data Feedback: At the end of the cycle, the synthesized data is fed back to the user, and adjustment suggestions are provided, including increasing rest time and adjusting diet.

[0051] It should be noted that: All calculation formulas in this application document use regression analysis including but not limited to machine learning algorithms to deeply analyze the relevant parameters collected and identify their natural trends and relationships. Use professional software, such as Python's Scikit-learn library or R language, to automatically generate mathematical models that match the data. Then, objectively evaluate the performance of the model through methods such as cross-validation, and combine continuous feedback and optimization to ensure that the created formula truly reflects the inherent laws of the data, thereby ensuring its effectiveness and accuracy. In all calculation formulas in this application, the parameters in each formula are dimensionlessly processed in a consistent range to ensure that different physical quantities are compared on the same scale; dimensionless technical means include but are not limited to Min-Max Normalization and Z-Score normalization; The technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0052] The logic and / or steps represented in the flowcharts or otherwise described herein, which can be considered as a sequenced list of executable instructions for implementing the logical functions, can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0053] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0054] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for time series fusion and health status prediction of multi-dimensional physiological data, characterized by: The specific steps include: Step 1: Real-time collection of multi-dimensional physiological data of the user at different exercise states during the current exercise and rehabilitation phase, and analysis of the multi-dimensional physiological data at different exercise states to construct a parameter calibration model; the multi-dimensional physiological data includes fatigue-related indicators and stress-related indicators; The parameter calibration model is used to automatically calibrate multidimensional physiological data of different motion states; Step 2: Obtain multidimensional physiological data after automatic calibration, use a weighted fusion algorithm to synchronize time and compensate for errors in the multidimensional physiological data of different motion states, and generate physiological parameter sequences of different motion states; Step 3: Based on the physiological parameter sequences of different exercise states, the fatigue index and stress index under different exercise states are calculated respectively, and closed-loop correction is performed; a threshold is preset, and the corrected fatigue index and stress index are compared and evaluated; Step 4: Based on the evaluation contents of fatigue index and stress index, further generate comprehensive state index under different exercise states; At the same time, the comprehensive state indexes under adjacent motion states were analyzed to generate a difference index; Step 5: Based on the difference index, predict the changing trend of the comprehensive status index within a preset time in the future; set a feedback cycle, and regularly provide advance feedback on the changing trend of the comprehensive status index to the user through wearable devices, and at the same time issue dynamic rehabilitation plan adjustment suggestions.

2. The method for time series fusion of multidimensional physiological data and health status prediction according to claim 1, characterized in that: Step 1 specifically includes: Wearable devices are configured to collect multi-dimensional physiological data in real time, including heart rate (HR), respiratory rate (BR), blood oxygen saturation (Xyb), heart rate variability (HRV), galvanic skin response (GSR), and cortisol level (Cor). These data are then transmitted to the wearable device's main control unit via BLE communication for unified processing. Timestamps are added to ensure that the multi-dimensional physiological data has temporal correlation. Based on the motion analysis results of the three-axis accelerometer in the wearable device, the different consecutive motion states in the current exercise and rehabilitation stage are sequentially marked to form a marking sequence {1, 2, ..., i, ..., n}, where i represents the sequential mark of the motion state at the current moment, and n represents the total number of motion states finally identified in the entire exercise and rehabilitation stage; According to the tag sequence, the collected multidimensional physiological data are associated with the corresponding motion state to form a multidimensional data set.

3. The method for time series fusion of multidimensional physiological data and health status prediction according to claim 2, characterized in that: Step 1 specifically includes: During the user's exercise, the main control MCU reads the original multi-dimensional physiological data currently marked as i in real time and outputs the calibration value through the trained vector regression model; If a change in the label sequence is detected, that is, the sequential label of the motion state switches from i to i+1, the online fine-tuning mechanism is triggered, and incremental learning is performed using real-time physiological data and the label sequence to update the model parameters; The calibrated multidimensional physiological data are transmitted to the application in real time via wireless communication.

4. The method for time series fusion of multidimensional physiological data and health status prediction according to claim 3, characterized in that: Step 2 specifically includes: The weighted fusion algorithm is used to perform error compensation and time synchronization processing on the multi-dimensional physiological data after automatic calibration; a continuous physiological parameter sequence {P1, P2, ..., P i ,…,P n }, each value represents the fusion physiological parameter value under a motion state, where P i represents the physiological parameter value of the i-th motion state, P n The physiological parameter value representing the nth motion state; Synchronize data with inconsistent timestamps by using linear interpolation, including calculating interpolated heart rate data; Error compensation is performed on the multidimensional physiological data after automatic calibration. The error calculation is based on the statistical difference of historical data, and dynamic error correction is performed using the Kalman filter; the Kalman filter dynamically adjusts the error estimate according to the difference between the real-time observation value and the predicted value.

5. The method for time series fusion of multi-dimensional physiological data and health status prediction according to claim 4, characterized in that: Step three specifically includes: Based on the heart rate, respiratory rate and blood oxygen saturation in the physiological parameter sequence, after standardization, {HR i ,BR i ,Xyb i } respectively represent the heart rate, respiratory rate and blood oxygen saturation in the i-th exercise state. The fatigue index is calculated using the following formula: in, is the fatigue index under the i-th exercise state, k1, k2 and k3 correspond to the weight coefficients of heart rate, respiratory rate and blood oxygen saturation respectively, and k1+k2+k3=1; Based on the heart rate variability, skin galvanic response and cortisol level in the physiological parameter sequence, standardization is performed and {HRV i ,GSR i ,Cor i } respectively represent the heart rate variability, skin galvanic response and cortisol level in the i-th exercise state. The stress index is calculated using the following formula: in, is the stress index under the i-th exercise state, k4, k5 and k6 are the weight coefficients of the corresponding heart rate variability, skin electrode response and cortisol level, and k4+k5+k6=1.

6. The method for time series fusion of multi-dimensional physiological data and health status prediction according to claim 5, characterized in that: Step three specifically includes: Perform real-time closed-loop feedback correction on fatigue index and stress index, extract two consecutive timestamps t from the data stream The fatigue index and pressure index corresponding to 1 and t are obtained by difference calculation to obtain fatigue deviation ΔApl and pressure deviation ΔAyl respectively; The specific content of the closed-loop feedback adjustment includes comprehensive evaluation and visual output of the corrected fatigue index and stress index; the evaluation logic is as follows: Set thresholds for fatigue index and stress index respectively; Normal range: If both the fatigue index and stress index are below their respective thresholds, it is considered normal and the user can continue with the current activity. Mild Abnormality: If either the fatigue index or the stress index exceeds the corresponding threshold, a mild abnormality warning will be issued and suggestions for adjusting activities will be made; Severe abnormality: If both the fatigue index and stress index exceed the corresponding thresholds, it is marked as a severe abnormality. At this time, the user is advised to adjust the exercise intensity in time.

7. The method for time series fusion of multi-dimensional physiological data and health status prediction according to claim 6, characterized in that: Step 4 specifically includes: The fatigue index and stress index are integrated to generate a comprehensive status index. The specific calculation formula is as follows: in, represents the comprehensive state index under the current motion state i, k7 and k8 are the weight parameters of the fusion fatigue index and pressure index respectively, and k7+k8=1; After generating the comprehensive state index, the difference index under adjacent motion states is analyzed by comparing the difference in the comprehensive state index between adjacent motion states i and i+1, and generating the difference index , the specific calculation formula is as follows: Difference Index Reflects the magnitude of change in stress and fatigue from the current state to the next state.

8. The method for time series fusion of multi-dimensional physiological data and health status prediction according to claim 7, characterized in that: Step 5 specifically includes: The comprehensive state index of the above motion state is established by machine learning method and difference index As input, a state prediction model is used to predict the trend of the comprehensive state index in the future; Set the timing of data synchronization and feedback cycles based on the user's specific needs and activity types; At the end of each cycle, the prediction result of the comprehensive state index is transmitted to the data transmission module on the user's wearable device; Among them, the prediction techniques of machine learning methods include linear regression and time series analysis.

9. The method for time series fusion of multi-dimensional physiological data and health status prediction according to claim 8, characterized in that: Step 5 specifically includes: Automatically generate personalized rehabilitation or training adjustment suggestions based on the predicted trend of the comprehensive status index. By analyzing each user's historical health data, current health status, and predicted future health trends, targeted suggestions are generated based on this analysis, including adjusting training intensity, optimizing diet plans, and adjusting rest periods. In addition, a user feedback mechanism is established to allow users to provide feedback on the suggestions received, which can be used to optimize the prediction model and adjust future health management strategies.

10. A multi-dimensional physiological data time series fusion and health status prediction system, characterized by: The system is used to execute the method for time series fusion and health status prediction of multi-dimensional physiological data according to any one of claims 1 to 9, comprising: The physiological data acquisition module collects multi-dimensional physiological data of the user in different exercise states in real time during the current exercise and rehabilitation stage, and analyzes the multi-dimensional physiological data of different exercise states to build a parameter calibration model; the multi-dimensional physiological data includes fatigue-related indicators and stress-related indicators; The parameter calibration model is used to automatically calibrate multidimensional physiological data of different motion states; The data fusion and calibration module acquires multi-dimensional physiological data after automatic calibration, uses a weighted fusion algorithm to synchronize time and compensate for errors in the multi-dimensional physiological data of different motion states, and generates physiological parameter sequences of different motion states; The index calculation and correction module calculates the fatigue index and stress index under different exercise states based on the physiological parameter sequences of different exercise states, and performs closed-loop correction. It presets thresholds and compares and evaluates the corrected fatigue index and stress index. The state index analysis module further generates comprehensive state indices under different exercise states based on the evaluation content of fatigue index and stress index; At the same time, the comprehensive state indexes under adjacent motion states were analyzed to generate a difference index; The trend prediction and feedback module predicts the changing trends of the comprehensive status index for multiple times within a preset time in the future based on the difference index; and sets a feedback cycle to regularly provide advance feedback on the changing trends of the comprehensive status index to users through wearable devices, while also issuing dynamic rehabilitation plan adjustment suggestions.

Images