Health risk assessment method and system based on multi-source data fusion analysis
By integrating and analyzing multi-source data and using smart wearable devices, and leveraging support vector machines and LSTM models, we have achieved real-time dynamic assessment and personalized recommendations for users' health status. This addresses the issues of real-time performance and individual variability adaptation in existing cerebrovascular disease assessments, thereby improving the effectiveness of health management.
Patent Information
- Application Number
- CN202410974411.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-19
- Publication Date
- 2024-11-12
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing methods for assessing the risk of cerebrovascular diseases lack real-time and dynamic capabilities, failing to reflect changes in users' health status in a timely manner. Furthermore, health management recommendations are based on static data and fixed thresholds, lacking adaptability to individual differences and thus reducing the effectiveness of the assessment.
By fusing and analyzing multi-source data, user behavior and physical data are collected. Support Vector Machine (SVM) is used for behavior classification, and LSTM model is combined to capture long-term dependencies in time series data. Health risk scores are calculated, and personalized exercise suggestions are generated. The data is securely uploaded to the cloud via wireless transmission to generate reports.
It enables precise monitoring of users' health status, provides real-time dynamic health risk assessment, predicts user behavior, and offers personalized exercise suggestions, thereby improving the effectiveness of cerebrovascular disease prevention and the real-time nature of health management.
Smart Images

Figure CN118940212A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of cerebrovascular risk assessment, and in particular to a health risk assessment method and system based on multi-source data fusion analysis. Background Art
[0002] In the field of modern medicine and health management, cerebrovascular disease has become a major health issue of global concern due to its high disability and mortality rates. Traditional health assessment methods mainly rely on periodic physical examinations and doctors' clinical judgments, lacking continuous monitoring of individuals' daily behaviors and physiological states, and are difficult to provide real-time and dynamic health risk assessments. In recent years, with the advancement of smart wearable devices, Internet of Things technology, and big data analysis technology, users' behavioral data and physical data can be collected in real time, and users' behavioral patterns and physiological changes can be analyzed.
[0003] However, the existing cerebrovascular disease risk assessment methods still have some shortcomings. In the process of data processing and analysis, the existing assessment systems often lack consideration of real-time and dynamic aspects and cannot promptly reflect changes in the user's health status. In addition, although some systems can provide health management recommendations, these recommendations are usually based on static health data and fixed threshold settings. Especially in the prevention and management of cerebrovascular diseases, they lack the ability to adapt to individual differences and dynamic changes, which reduces the risk assessment effect of the data. Summary of the invention
[0004] In view of the problems existing in the above-mentioned existing health risk assessment methods and systems based on multi-source data fusion analysis, the present invention is proposed.
[0005] Therefore, the problem to be solved by the present invention is that the existing evaluation systems often lack consideration of real-time and dynamic aspects during data processing and analysis, and cannot promptly reflect changes in the user's health status. In addition, although some systems can provide health management recommendations, these recommendations are usually based on static health data and fixed threshold settings. Especially in the prevention and management of cerebrovascular diseases, they lack the ability to adapt to individual differences and dynamic changes, thereby reducing the risk assessment effect of the data.
[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a health risk assessment method based on multi-source data fusion analysis, which comprises:
[0007] Collect user behavior data and physical data, and evaluate changes in user physical data;
[0008] According to the changes in physical data, the user behavior is analyzed through the support vector machine (SVM), and the health risk score is calculated and the user's health risk is determined through the user behavior analysis;
[0009] Predict user behavior based on user physical data, count user behavior time, calculate recommended exercise time based on user health risk, and conduct comprehensive risk assessment based on user behavior;
[0010] Data is securely transmitted to the cloud via wireless transmission, and reports are generated.
[0011] As a preferred solution of the health risk assessment method based on multi-source data fusion analysis of the present invention, wherein: the collection of user behavior data and physical data, and the assessment of changes in user physical data include:
[0012] Collect user behavior data and physical data based on smart wearable devices. Behavioral data includes GPS coordinates and timestamps, and physical data includes steps, heart rate, and blood pressure.
[0013] Calculate the user's BMR based on their basic information including gender, age, weight and height, and calculate calorie consumption with heart rate;
[0014] Perform data cleaning and standard deviation processing on the collected behavioral data and physical data, and unify the time format;
[0015] Calculate the user's activity frequency in a specific time interval, expressed as:
[0016]
[0017] where F s represents the activity frequency, Δt represents the time interval, N s Indicates the number of steps;
[0018] Calculate the change in the user's heart rate, expressed as:
[0019]
[0020] Where ΔHR represents the change in heart rate, HR max Indicates maximum heart rate, HR min Indicates minimum heart rate;
[0021] Calculate the range of the user's blood pressure, expressed as:
[0022]
[0023] ΔBP represents the change in blood pressure, BP max Indicates maximum blood pressure, BP min Indicates minimum blood pressure;
[0024] Calculate the frequency and range of the user's location change within the time interval and evaluate their movement, expressed as:
[0025]
[0026] where F l Indicates the frequency of position change, N l represents the number of position changes, R represents the maximum position change between point i and point j within the time interval Δt, x i and i represents the i-th GPS coordinate at the start time of the time interval Δt, x j and j Indicates the j-th GPS coordinate at the end of the time interval Δt.
[0027] As a preferred solution of the health risk assessment method based on multi-source data fusion analysis of the present invention, wherein: the user behavior is analyzed by support vector machine SVM according to the changes in body data, including:
[0028] Based on the activity frequency F s , heart rate change ΔHR, blood pressure change ΔBP, position change frequency F l And the position change value R constructs the feature vector X, which is expressed as:
[0029] X=[F s ,ΔHR,ΔBP,R,F l ];
[0030] Build a classification model and use the support vector machine (SVM) algorithm for behavior classification;
[0031] Use the training set to evaluate the model performance through cross-validation, use the Adam optimizer and the gradient descent method to iteratively optimize the model parameters. When the loss of the classification model no longer decreases significantly during the continuous iteration process, stop iterating and output the model parameters to update the classification model.
[0032] Collect data through smart wearable devices, build a feature vector input model, and classify user behaviors.
[0033] As a preferred solution of the health risk assessment method based on multi-source data fusion analysis of the present invention, wherein: the health risk scoring and judging the user's health risk through user behavior analysis includes:
[0034] Based on the different behavior classification results predicted by the classification model, the relationship between blood pressure changes, heart rate changes and activity frequency in the current behavior classification results is comprehensively considered to calculate the health risk score, which is expressed as:
[0035]
[0036] HRI stands for health risk score, BP b represents baseline blood pressure, HR brepresents baseline heart rate;
[0037] Based on statistical data, the HRI distribution of different behavioral classification results of users in a healthy state is calculated, and the value below the 25th percentile is taken as the low-risk threshold. The HRI distribution of the corresponding behavioral classification results of users in a cerebrovascular pathological state is calculated, and the value above the 75th percentile is taken as the high-risk threshold.
[0038] If the calculated HRI value is greater than or equal to the high-risk threshold, the user's current health status is judged to be high-risk;
[0039] If the calculated HRI value is greater than or equal to the low risk threshold and less than the high risk threshold, the user's current health status is judged to be medium risk;
[0040] If the calculated HRI value is less than the low risk threshold, the user's current health status is judged to be low risk.
[0041] As a preferred solution of the health risk assessment method based on multi-source data fusion analysis of the present invention, wherein: the prediction of user behavior based on user body data and the statistics of user behavior time include:
[0042] Based on the activity frequency F s , heart rate change ΔHR, blood pressure change ΔBP, position change frequency F l The time series data of the position change value R is used as the feature value to form the feature vector H, and the LSTM model is constructed to capture the long-term dependency in the time series data;
[0043] The model is used to predict the user behavior classification at future time points, and the total time of the user's non-sports behavior classification and the total time of the sports behavior classification are counted respectively.
[0044] As a preferred solution of the health risk assessment method based on multi-source data fusion analysis of the present invention, wherein: the reference user health risk calculation recommended exercise time, and the comprehensive risk assessment based on the user behavior include:
[0045] Total time T based on user non-sports behavior classification n and the total time T for user motion behavior classification o Calorie expenditure is calculated based on the total calorie expenditure, expressed as:
[0046] C p =BMR·T n +(me±std)·T o ;
[0047] Where me and std represent the mean and standard deviation of the exercise calorie consumption rate in the historical data, respectively. pIndicates the user's total calorie consumption, and BMR indicates the user's basal metabolic rate;
[0048] Set the recommended calorie consumption Z for low-risk health state, the recommended calorie consumption B for medium-risk health state, and the recommended calorie consumption G for high-risk health state;
[0049] The exercise type recommendation for low-risk health status and medium-risk health status is set as moderate-intensity exercise, and the exercise type recommendation for high-risk health status is set as low-intensity exercise;
[0050] According to the standard of energy expenditure reference value of physical activity for healthy adults, the metabolic calorie expenditure of moderate-intensity exercise and low-intensity exercise is determined;
[0051] Calculate the recommended exercise time based on the recommended calorie expenditure, expressed as:
[0052]
[0053] Where K represents the user's recommended exercise time, including the recommended exercise time for low-risk health status, the recommended exercise time for medium-risk health status, and the recommended exercise time for high-risk health status;
[0054] G p Indicates the total calorie consumption of the user, including the total calorie consumption of a low-risk health state, the total calorie consumption of a medium-risk health state, and the total calorie consumption of a high-risk health state;
[0055] C u Indicates the user's metabolic calorie consumption rate, including the metabolic calorie consumption rate of moderate-intensity exercise and the metabolic calorie consumption rate of high-intensity exercise;
[0056] L stands for recommended calorie consumption, including recommended calorie consumption for low-risk health status, recommended calorie consumption for medium-risk health status, and recommended calorie consumption for high-risk health status;
[0057] Based on the actual calorie consumption data collected by the smart wearable device, calculate the achievement rate A of the recommended calorie consumption;
[0058] The dynamic changes of user's body data based on the analysis of user data are expressed as:
[0059]
[0060] Where DAF represents the dynamic adjustment factor, ΔHR a Indicates the average heart rate change, ΔBP a represents the mean blood pressure change, σ HR represents the standard deviation of heart rate variation, σ BP represents the standard deviation of blood pressure change;
[0061] A comprehensive risk assessment is performed based on the user's calorie consumption achievement rate, expressed as:
[0062] CHR=(HRI·DAF·LF)·(1-A)+δ·A;
[0063] Where LF represents the lifestyle factor, and δ represents the influence coefficient of the achievement rate A;
[0064] Based on statistical data, the CHR distribution of users in a healthy state under different behavior classification results is calculated, and the value below the 50th percentile of the distribution is taken as the comprehensive threshold. If the CHR is greater than the comprehensive threshold, the comprehensive risk is considered to be high; if the CHR is less than or equal to the comprehensive threshold, the comprehensive risk is considered to be low.
[0065] As a preferred solution of the health risk assessment method based on multi-source data fusion analysis described in the present invention, wherein: the data is securely transmitted to the cloud through wireless transmission, and a report is generated, including:
[0066] The smart wearable device regularly uploads data to the cloud server through the secure communication protocol TLS based on the wireless network;
[0067] Set up a data receiving service in the cloud and use the reporting tool Power BI to automatically generate reports. Use the scheduling function of the cloud service to regularly generate and update reports. The reports include collected user behavior data and physical data, behavior classification results, health risk scores, users' total calorie consumption, and recommended exercise time.
[0068] Another object of the present invention is to provide a system for health risk assessment method based on multi-source data fusion analysis, which comprises:
[0069] Data collection module, collects user behavior data and physical data in real time;
[0070] The behavior classification module analyzes the collected data and classifies the behavior through support vector machine (SVM);
[0071] The health risk assessment module performs health risk scoring and determines the user's health risk based on the behavior classification results and physical data changes;
[0072] Behavior prediction module, which predicts user behavior based on user physical data and counts the time of different behavior patterns;
[0073] The exercise suggestion module calculates calorie consumption based on user behavior time, and calculates the recommended exercise time based on the health risk score, and conducts a comprehensive risk assessment based on user behavior;
[0074] The cloud processing module securely transmits data to the cloud via wireless transmission and uses reporting tools to automatically generate and update health reports.
[0075] A computer device comprises: a memory and a processor; the memory stores a computer program, and the processor implements the steps of the above-mentioned health risk assessment method based on multi-source data fusion analysis when executing the computer program.
[0076] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned health risk assessment method based on multi-source data fusion analysis.
[0077] The beneficial effects of the present invention are as follows: through multi-dimensional data collection and analysis, accurate monitoring of user health conditions can be achieved; through the calculation of user health risk scores, real-time dynamic health risk assessment is provided for users with cerebrovascular diseases; the long-term dependencies of time series data are captured by the LSTM model, and the user's behavior classification can be accurately predicted, which helps to understand the user's daily activity patterns and further provide data support for health management; dynamic, scientific and accurate exercise suggestions are provided based on user health risks and behavior classifications; users are helped to achieve health goals based on the physical data of different users; feedback and comprehensive evaluation of users' actual behaviors are performed, achieving real-time and dynamic considerations based on user behaviors to prevent the occurrence of cerebrovascular diseases. BRIEF DESCRIPTION OF THE DRAWINGS
[0078] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.
[0079] Figure 1 This is a flowchart of the health risk assessment method and system based on multi-source data fusion analysis.
[0080] Figure 2 This is a schematic diagram of the structure of the health risk assessment method and system based on multi-source data fusion analysis. DETAILED DESCRIPTION
[0081] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.
[0082] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.
[0083] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.
[0084] Example 1, reference Figure 1 , which is the first embodiment of the present invention, and which provides a health risk assessment method based on multi-source data fusion analysis. The health risk assessment method based on multi-source data fusion analysis includes:
[0085] S1, collect user behavior data and physical data, and evaluate changes in user physical data;
[0086] Preferably, collecting user behavior data and physical data, and evaluating changes in user physical data, includes:
[0087] Collect user behavior data and physical data based on smart wearable devices. Behavioral data includes GPS coordinates and timestamps, and physical data includes steps, heart rate, and blood pressure.
[0088] Calculate the user's BMR based on their basic information including gender, age, weight and height, and calculate calorie consumption with heart rate;
[0089] Perform data cleaning and standard deviation processing on the collected behavioral data and physical data, and unify the time format;
[0090] Calculate the user's activity frequency in a specific time interval, expressed as:
[0091]
[0092] where F s represents the activity frequency, Δt represents the time interval, N s Indicates the number of steps;
[0093] Calculate the change in the user's heart rate, expressed as:
[0094]
[0095] Where ΔHR represents the change in heart rate, HR max Indicates maximum heart rate, HR min Indicates minimum heart rate;
[0096] Calculate the range of the user's blood pressure, expressed as:
[0097]
[0098] ΔBP represents the change in blood pressure, BP max Indicates maximum blood pressure, BP min Indicates minimum blood pressure;
[0099] Calculate the frequency and range of the user's location change within the time interval and evaluate their movement, expressed as:
[0100]
[0101] where F l Indicates the frequency of position change, N l represents the number of position changes, R represents the maximum position change between point i and point j within the time interval Δt, x i and i represents the i-th GPS coordinate at the start time of the time interval Δt, x j and j Indicates the j-th GPS coordinate at the end of the time interval Δt.
[0102] Through multi-dimensional data collection and analysis, accurate monitoring of user health status can be achieved. According to changes in user behavior and physical data, especially the risk of cerebrovascular disease, this helps to detect health problems early and provide timely intervention measures. Personalized health advice can be provided at one time to improve the effectiveness of user health management. By analyzing user behavior data and identifying their life and exercise patterns, a basis for further health analysis and advice can be provided. Through continuous data collection and analysis, users' long-term health status can be tracked to help users better understand and manage their own health. In particular, monitoring based on blood pressure and heart rate data can be viewed by users at all times to prevent the occurrence and trends of cerebrovascular disease and related complications, thereby achieving risk prevention effects.
[0103] S2, according to the changes in body data, analyzes user behavior through support vector machine (SVM), and uses user behavior analysis to score health risks and determine user health risks;
[0104] Preferably, according to the changes in body data, the user behavior is analyzed by a support vector machine (SVM), including:
[0105] Based on the activity frequency F s , heart rate change ΔHR, blood pressure change ΔBP, position change frequency F l And the position change value R constructs the feature vector X, which is expressed as:
[0106] X=[F s ,ΔHR,ΔBP,R,F l ];
[0107] Construct a classification model and select the support vector machine (SVM) algorithm for behavior classification, which is expressed as:
[0108] y = f(X);
[0109] Where y represents the behavior classification result, and y can be classified into work mode, rest mode and exercise mode based on user habits. X represents the feature vector, and f represents the SVM classifier.
[0110] The training set is used to evaluate the model performance through cross-validation to ensure the classification accuracy, which is expressed as:
[0111]
[0112] Where ACC represents the model accuracy, N represents the total number of samples, and y i represents the true label of the i-th sample, represents the predicted label of the i-th sample, If the true label and the predicted label are equal, the value is 1, indicating a correct prediction; if they are not equal, the value is 0, indicating an incorrect prediction.
[0113] The Adam optimizer is used to iteratively optimize the model parameters through the gradient descent method. When the loss of the classification model no longer decreases significantly during the continuous iteration process, the iteration is stopped and the model parameters are output to update the classification model.
[0114] Collect data through smart wearable devices, build a feature vector input model, and classify user behaviors.
[0115] By utilizing the real-time data collection function of smart wearable devices and analyzing user behavior patterns and health data, we can provide users with personalized health management plans and provide data support for subsequent exercise recommendations and health risk assessments. Based on the identification of user behavior classification, we can generate intelligent health recommendations, including diet, exercise, and rest, to help users improve their health overall and provide data support for the health management functions of smart wearable devices.
[0116] Furthermore, health risk scores are scored and user health risks are determined through user behavior analysis, including:
[0117] Based on the different behavior classification results predicted by the classification model, the relationship between blood pressure changes, heart rate changes and activity frequency in the current behavior classification results is comprehensively considered to calculate the health risk score, which is expressed as:
[0118]
[0119] HRI stands for health risk score, BP b represents baseline blood pressure, HR b It indicates baseline heart rate. Baseline blood pressure and baseline heart rate can be determined based on the guidelines on hypertension and heart rate issued by the World Health Organization, hypertension health management standards, and multidisciplinary expert consensus on heart rate management, or can be given by a doctor based on the patient's cerebrovascular disease.
[0120] Based on statistical data, the HRI distribution of different behavioral classification results of users in a healthy state is calculated, and the value below the 25th percentile is taken as the low-risk threshold. The HRI distribution of the corresponding behavioral classification results of users in a cerebrovascular pathological state is calculated, and the value above the 75th percentile is taken as the high-risk threshold.
[0121] If the calculated HRI value is greater than or equal to the high-risk threshold, the user's current health status is judged to be high-risk;
[0122] If the calculated HRI value is greater than or equal to the low risk threshold and less than the high risk threshold, the user's current health status is judged to be medium risk;
[0123] If the calculated HRI value is less than the low risk threshold, the user's current health status is judged to be low risk.
[0124] By calculating the user's health risk score, focusing on data related to heart rate and blood pressure as well as baseline data related to experts and medical organizations, a special health risk assessment is provided for users with cerebrovascular diseases. Through numerical assessment calculations, patients with cerebrovascular diseases are provided with intuitive numerical values to represent the risk index of cerebrovascular-related diseases. At the same time, risk judgment based on the health risk score can provide users with risk warnings more intuitively, allowing users to go to the hospital for consultation and treatment in time when the risk is medium or high, and provide data support for subsequent exercise recommendations.
[0125] S3, predicts user behavior based on user physical data, counts user behavior time, calculates recommended exercise time based on user health risk, and conducts comprehensive risk assessment based on user behavior;
[0126] Preferably, predicting user behavior based on user body data and counting user behavior time include:
[0127] Based on the activity frequency F s , heart rate change ΔHR, blood pressure change ΔBP, position change frequency F l The time series data of the position change value R is used as the feature value to form the feature vector H. The LSTM model is constructed to capture the long-term dependency in the time series data, which is expressed as:
[0128]
[0129] in represents the model prediction for the tth time interval, W y represents the weight of the output layer, b y represents the bias of the output layer, h t represents the eigenvalue of the tth time interval;
[0130] Use the training set to train the model and define the mean square error loss function, which is expressed as:
[0131]
[0132] Where M represents the total number of samples, u i represents the actual behavior state of the i-th sample, represents the predicted behavior state of the i-th sample;
[0133] The Adam optimizer uses the back propagation algorithm to minimize the loss function and update the model's weights and bias parameters using the gradient descent method. When the loss of the LSTM model no longer decreases significantly during continuous iterations, the iteration is stopped and the model parameters are output to update the LSTM model.
[0134] The model is used to predict the user behavior classification at future time points, and the total time of the user's non-sports behavior classification and the total time of the sports behavior classification are counted respectively, which is expressed as:
[0135]
[0136] Where T is the total number of time intervals, T n Represents the total time of user non-sports behavior classification, T o Indicates the total time of user motion behavior classification.
[0137] By capturing the long-term dependencies of time series data through the LSTM model, the user's behavior classification can be accurately predicted. Based on the time statistics of feature vectors and behavior classification, the system can monitor the user's health status in real time. This precise classification helps to understand the user's daily activity patterns and further provide data support for health management. Through the predicted user behavior and the statistical user behavior time, it can be preliminarily determined whether the user is in different states such as work, rest or exercise, and it helps to understand the user's living habits, so as to predict the user's exercise consumption according to the user's behavior state time, and give exercise suggestions in advance to prevent further severity of cerebrovascular diseases.
[0138] Furthermore, the recommended exercise time is calculated based on the user's health risk, and a comprehensive risk assessment is performed based on the user's behavior, including:
[0139] Total time T based on user non-sports behavior classification n and the total time T for user motion behavior classification o Calorie expenditure is calculated based on the total calorie expenditure, expressed as:
[0140] C p =BMR·T n +(me±std)·T o ;
[0141] Where me and std represent the mean and standard deviation of the exercise calorie consumption rate in the historical data, respectively. p Indicates the user's total calorie consumption, and BMR indicates the user's basal metabolic rate;
[0142] Based on the determination of health risk scores, and in accordance with the guidelines on hypertension and heart rate from the World Health Organization, the Hypertension Health Management Standards, and the Multidisciplinary Expert Consensus on Heart Rate Management;
[0143] Set the recommended calorie consumption Z for low-risk health state, the recommended calorie consumption B for medium-risk health state, and the recommended calorie consumption G for high-risk health state;
[0144] The exercise type recommendation for low-risk health status and medium-risk health status is set as moderate-intensity exercise, and the exercise type recommendation for high-risk health status is set as low-intensity exercise;
[0145] According to the standard of energy expenditure reference value of physical activity for healthy adults, the metabolic calorie expenditure of moderate-intensity exercise and low-intensity exercise is determined, expressed as:
[0146] C g =M g ·Z;
[0147] C d =M d ·Z;
[0148] Among them C g Indicates the metabolic calorie consumption rate of moderate intensity exercise, C d Indicates the metabolic calorie consumption rate of low-intensity exercise, M g The MET value represents the metabolic equivalent of moderate intensity exercise. d It represents the metabolic equivalent MET value of low-intensity exercise, and Z represents the user's weight;
[0149] Calculate the recommended exercise time based on the recommended calorie expenditure, expressed as:
[0150]
[0151] Where K represents the user's recommended exercise time, including the recommended exercise time for low-risk health status, the recommended exercise time for medium-risk health status, and the recommended exercise time for high-risk health status;
[0152] G p Indicates the total calorie consumption of the user, including the total calorie consumption of a low-risk health state, the total calorie consumption of a medium-risk health state, and the total calorie consumption of a high-risk health state;
[0153] C u Indicates the user's metabolic calorie consumption rate, including the metabolic calorie consumption rate of moderate-intensity exercise and the metabolic calorie consumption rate of high-intensity exercise;
[0154] L stands for recommended calorie consumption, including recommended calorie consumption for low-risk health status, recommended calorie consumption for medium-risk health status, and recommended calorie consumption for high-risk health status;
[0155] Based on the actual calorie consumption data collected by the smart wearable device, calculate the achievement rate A of the recommended calorie consumption;
[0156] The dynamic changes of user's body data based on the analysis of user data are expressed as:
[0157]
[0158] Where DAF represents the dynamic adjustment factor, ΔHR a Indicates the average heart rate change, ΔBP a represents the mean blood pressure change, σ HR represents the standard deviation of heart rate variation, σ BP represents the standard deviation of blood pressure change;
[0159] A comprehensive risk assessment is performed based on the user's calorie consumption achievement rate, expressed as:
[0160] CHR=(HRI·DAF·LF)·(1-A)+δ·A;
[0161] Where LF represents the lifestyle factor (calorie consumption classified by non-exercise behavior), δ represents the influence coefficient of the achievement rate A;
[0162] The linear regression model is used to evaluate the impact of the achievement rate A on the comprehensive risk assessment, which is expressed as:
[0163] CHR'=β0+δ·A+∈;
[0164] Where CHR' represents the model output, β0 represents the model intercept, and ∈ represents the model error term;
[0165] Using the training set and the mean square error loss function, it is expressed as:
[0166]
[0167] Where O represents the calculation loss, P represents the total number of samples, and CHR i represents the actual value of the i-th CHR;
[0168] The model parameters are iteratively optimized using the Adam optimizer and the gradient descent method. When the loss of the linear regression model no longer decreases significantly during the continuous iteration, the iteration is stopped and the model parameters are output to update the linear regression model and determine the influence coefficient δ of the achievement rate A.
[0169] Based on statistical data, the CHR distribution of users in a healthy state under different behavior classification results is calculated, and the value below the 50th percentile of the distribution is taken as the comprehensive threshold. If the CHR is greater than the comprehensive threshold, the comprehensive risk is considered to be high; if the CHR is less than or equal to the comprehensive threshold, the comprehensive risk is considered to be low.
[0170] By calculating the user's total calorie consumption and health risk score, the system can provide users with scientific and accurate exercise recommendations. The system provides personalized exercise recommendations based on the user's health risk score and current calorie consumption, including specific exercise types and exercise time. This personalized recommendation can effectively help users achieve their health goals and prevent the occurrence of cerebrovascular diseases. For high-risk users, the system provides low-intensity exercise recommendations to ensure that users exercise within a safe range and reduce the risk of sudden cerebrovascular diseases. For medium-risk and low-risk users, the system provides moderate-intensity exercise recommendations to help users gradually improve their cardiopulmonary function, improve blood pressure and heart rate conditions, and prevent the occurrence of cerebrovascular diseases. At the same time, the calculation of the achievement rate of the recommended exercise time can effectively feedback and record the user's personal exercise completion, and give a comprehensive judgment based on this, and give a comprehensive risk assessment of the user's actual behavior. Through scientific exercise recommendations, users can form healthy living habits and improve their overall quality of life. The system's real-time monitoring and feedback can help users maintain a good health state and reduce the probability of disease. At the same time, the system can also record the user's actual exercise calorie consumption achievement rate based on smart wearable devices, thereby further evaluating the user's health risks and providing users with further risk warnings.
[0171] S4, securely transmits data to the cloud via wireless transmission and generates reports;
[0172] Preferably, the data is securely transmitted to the cloud via wireless transmission, and a report is generated, including:
[0173] The smart wearable device regularly uploads data to the cloud server through the secure communication protocol TLS based on the wireless network;
[0174] Set up a data receiving service in the cloud and use the reporting tool Power BI to automatically generate reports. Use the scheduling function of the cloud service to regularly generate and update reports. The reports include collected user behavior data and physical data, behavior classification results, health risk scores, users' total calorie consumption, and recommended exercise time.
[0175] By using the TLS protocol to transmit data, the security of data during transmission is ensured to prevent data from being stolen or tampered with. This security mechanism provides users with a high degree of privacy protection and enhances users' trust in the system. Smart wearable devices collect users' behavior and physical data in real time, and upload them to the cloud server in a timely manner via wireless networks to achieve real-time data updates and processing. Users can understand their health status and exercise recommendations at any time through different devices connected to the cloud, ensuring the real-time and accuracy of health management. Users can also understand their health status and exercise status in detail through report generation, which has the effect of intuitively understanding body data.
[0176] Example 2, reference Figure 2 , which is the second embodiment of the present invention, and which is different from the previous embodiment, provides a system for health risk assessment method based on multi-source data fusion analysis, including:
[0177] Data collection module, collects user behavior data and physical data in real time;
[0178] The behavior classification module analyzes the collected data and classifies the behavior through support vector machine (SVM);
[0179] The health risk assessment module performs health risk scoring and determines the user's health risk based on the behavior classification results and physical data changes;
[0180] Behavior prediction module, which predicts user behavior based on user physical data and counts the time of different behavior patterns;
[0181] The exercise suggestion module calculates calorie consumption based on user behavior time, and calculates the recommended exercise time based on the health risk score, and conducts a comprehensive risk assessment based on user behavior;
[0182] The cloud processing module securely transmits data to the cloud via wireless transmission and uses reporting tools to automatically generate and update health reports.
[0183] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.
[0184] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.
[0185] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.
[0186] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0187] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical 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 health risk assessment method based on multi-source data fusion analysis, characterized by: include, Collect user behavior data and physical data, and evaluate changes in user physical data; According to the changes in physical data, the user behavior is analyzed through the support vector machine (SVM), and the health risk score is calculated and the user's health risk is determined through the user behavior analysis; Predict user behavior based on user physical data, count user behavior time, calculate recommended exercise time based on user health risk, and conduct comprehensive risk assessment based on user behavior; Data is securely transmitted to the cloud via wireless transmission, and reports are generated.
2. The health risk assessment method based on multi-source data fusion analysis according to claim 1, characterized in that: The collecting of user behavior data and physical data and evaluating changes in user physical data include: Collect user behavior data and physical data based on smart wearable devices. Behavioral data includes GPS coordinates and timestamps, and physical data includes steps, heart rate, and blood pressure. Calculate the user's BMR based on their basic information including gender, age, weight and height, and calculate calorie consumption with heart rate; Perform data cleaning and standard deviation processing on the collected behavioral data and physical data, and unify the time format; Calculate the user's activity frequency within a specific time interval, expressed as: Among them, F s represents the activity frequency, Δt represents the time interval, N s Indicates the number of steps; Calculate the change in the user's heart rate, expressed as: Where ΔHR represents the change in heart rate, HR max Indicates maximum heart rate, HR min Indicates minimum heart rate; Calculate the range of the user's blood pressure, expressed as: ΔBP represents the change in blood pressure, BP max Indicates maximum blood pressure, BP min Indicates minimum blood pressure; Calculate the frequency and range of the user's location change within the time interval and evaluate their movement, expressed as: where F l Indicates the frequency of position change, N l represents the number of position changes, R represents the maximum position change between point i and point j within the time interval Δt, x i and i represents the i-th GPS coordinate at the start time of the time interval Δt, x j and j Indicates the j-th GPS coordinate at the end of the time interval Δt.
3. The health risk assessment method based on multi-source data fusion analysis as claimed in claim 2, characterized in that: According to the changes in body data, the user behavior is analyzed by support vector machine SVM. include, Based on the activity frequency F s , heart rate change ΔHR, blood pressure change ΔBP, position change frequency F l And the position change value R constructs the feature vector X, which is expressed as: X=[F s ,ΔHR,ΔBP,R,F l ]; Build a classification model and use the support vector machine (SVM) algorithm for behavior classification; Use the training set to evaluate the model performance through cross-validation, use the Adam optimizer and the gradient descent method to iteratively optimize the model parameters. When the loss of the classification model no longer decreases significantly during the continuous iteration process, stop iterating and output the model parameters to update the classification model. Collect data through smart wearable devices, build a feature vector input model, and classify user behaviors.
4. The health risk assessment method based on multi-source data fusion analysis as claimed in claim 3, characterized in that: The health risk scoring and judging the user's health risk through user behavior analysis includes: Based on the different behavior classification results predicted by the classification model, the relationship between blood pressure changes, heart rate changes and activity frequency in the current behavior classification results is comprehensively considered to calculate the health risk score, which is expressed as: HRI stands for health risk score, BP b represents baseline blood pressure, HR b represents baseline heart rate; Based on statistical data, the HRI distribution of different behavioral classification results of users in a healthy state is calculated, and the value below the 25th percentile is taken as the low-risk threshold. The HRI distribution of the corresponding behavioral classification results of users in a cerebrovascular pathological state is calculated, and the value above the 75th percentile is taken as the high-risk threshold. If the calculated HRI value is greater than or equal to the high-risk threshold, the user's current health status is judged to be high-risk; If the calculated HRI value is greater than or equal to the low risk threshold and less than the high risk threshold, the user's current health status is judged to be medium risk; If the calculated HRI value is less than the low risk threshold, the user's current health status is judged to be low risk.
5. The health risk assessment method based on multi-source data fusion analysis as claimed in claim 4, characterized in that: The method predicts user behavior based on user body data and counts user behavior time. include, Based on the activity frequency F s , heart rate change ΔHR, blood pressure change ΔBP, position change frequency F l The time series data of the position change value R is used as the feature value to form the feature vector H, and the LSTM model is constructed to capture the long-term dependency in the time series data; The model is used to predict the user behavior classification at future time points, and the total time of the user's non-sports behavior classification and the total time of the sports behavior classification are counted respectively.
6. The health risk assessment method based on multi-source data fusion analysis as claimed in claim 5, characterized in that: The reference user health risk calculation recommends exercise time and conducts a comprehensive risk assessment based on user behavior. include, Total time T based on user non-sports behavior classification n and the total time T for user motion behavior classification o Calorie expenditure is calculated based on the total calorie expenditure, expressed as: C p =BMR·T n +(me±std)·T o ; Where me and std represent the mean and standard deviation of the exercise calorie consumption rate in the historical data, respectively. p Indicates the user's total calorie consumption, and BMR indicates the user's basal metabolic rate; Set the recommended calorie consumption Z for low-risk health state, the recommended calorie consumption B for medium-risk health state, and the recommended calorie consumption G for high-risk health state; The exercise type recommendation for low-risk health status and medium-risk health status is set as moderate-intensity exercise, and the exercise type recommendation for high-risk health status is set as low-intensity exercise; According to the standard of energy expenditure reference value of physical activity for healthy adults, the metabolic calorie expenditure of moderate-intensity exercise and low-intensity exercise is determined; Calculate the recommended exercise time based on the recommended calorie expenditure, expressed as: Where K represents the user's recommended exercise time, including the recommended exercise time for low-risk health status, the recommended exercise time for medium-risk health status, and the recommended exercise time for high-risk health status; G p Indicates the total calorie consumption of the user, including the total calorie consumption of a low-risk health state, the total calorie consumption of a medium-risk health state, and the total calorie consumption of a high-risk health state; C u Indicates the user's metabolic calorie consumption rate, including the metabolic calorie consumption rate of moderate-intensity exercise and the metabolic calorie consumption rate of high-intensity exercise; L stands for recommended calorie consumption, including recommended calorie consumption for low-risk health status, recommended calorie consumption for medium-risk health status, and recommended calorie consumption for high-risk health status; Based on the actual calorie consumption data collected by the smart wearable device, calculate the achievement rate A of the recommended calorie consumption; The dynamic changes of user's body data based on the analysis of user data are expressed as: Where DAF represents the dynamic adjustment factor, ΔHR a Indicates the average heart rate change, ΔBP a represents the mean blood pressure change, σ HR represents the standard deviation of heart rate variation, σ BP represents the standard deviation of blood pressure change; A comprehensive risk assessment is performed based on the user's calorie consumption achievement rate, expressed as: CHR=(HRI·DAF·LF)·(1-A)+δ·A; Where LF represents the lifestyle factor, and δ represents the influence coefficient of the achievement rate A; Based on statistical data, the CHR distribution of users in a healthy state under different behavior classification results is calculated, and the value below the 50th percentile of the distribution is taken as the comprehensive threshold. If the CHR is greater than the comprehensive threshold, the comprehensive risk is considered to be high; if the CHR is less than or equal to the comprehensive threshold, the comprehensive risk is considered to be low.
7. The health risk assessment method based on multi-source data fusion analysis according to claim 6 is characterized in that: The data is securely transmitted to the cloud via wireless transmission, and a report is generated, including: The smart wearable device regularly uploads data to the cloud server through the secure communication protocol TLS based on the wireless network; Set up a data receiving service in the cloud and use the reporting tool Power BI to automatically generate reports. Use the scheduling function of the cloud service to regularly generate and update reports. The reports include collected user behavior data and physical data, behavior classification results, health risk scores, users' total calorie consumption, and recommended exercise time.
8. A system based on the health risk assessment method based on multi-source data fusion analysis according to any one of claims 1 to 7, characterized in that: include, Data collection module, collects user behavior data and physical data in real time; The behavior classification module analyzes the collected data and classifies the behavior through support vector machine (SVM); The health risk assessment module performs health risk scoring and determines the user's health risk based on the behavior classification results and physical data changes; Behavior prediction module, which predicts user behavior based on user physical data and counts the time of different behavior patterns; The exercise suggestion module calculates calorie consumption based on user behavior time, and calculates the recommended exercise time based on the health risk score, and conducts a comprehensive risk assessment based on user behavior; The cloud processing module transmits data securely to the cloud via wireless transmission and uses reporting tools to automatically generate and update health reports.
9. A computer device comprising: Memory and processor; The memory stores a computer program, characterized in that when the processor executes the computer program, the steps of the health risk assessment method based on multi-source data fusion analysis described in any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the health risk assessment method based on multi-source data fusion analysis described in any one of claims 1 to 7 are implemented.
Citation Information
Patent Citations
Exercise prescription recommendation system for old people
CN116705233A
Dynamic risk assessment method and system based on intelligent wearable device
CN117877732A
Monitoring patient's health
US11342076B1
Cited By
Intensive care unit patient early warning method based on clinical data
CN119153103A
Automatic input vital sign detection method and system
CN119324065A
Fatigue risk level evaluation system and device and storage medium
CN119361159A
Health risk assessment system based on multi-source data fusion analysis of intelligent wearable device
CN119538060A
Health risk assessment system based on multi-source data fusion analysis of smart wearable devices
CN119538060B
