Precise health index prediction method and system based on multi-source data fusion

Abstract

The invention discloses an accurate health index prediction method and system based on multi-source data fusion, and relates to the technical field of health management. The method comprises the following steps: acquiring multi-dimensional data of a user; according to the preprocessed multi-dimensional data, biological harmony characteristics are constructed, wherein the biological harmony characteristics comprise a physiological health comprehensive index, a psychological health fluctuation index, an activity intensity adaptation index and an environment stress index; calculating a comprehensive health dynamic index according to the biological harmony characteristics; performing threshold judgment on the comprehensive health dynamic index, and providing corresponding health suggestions; constructing an LSTM model based on the historical comprehensive health dynamic index to perform trend prediction; and the health state and the health trend prediction result are monitored in real time through a visual billboard. Comprehensive health assessment is performed by integrating multi-dimensional data, personalized health intervention is provided, and a health threshold is dynamically adjusted according to personal information of a user, so that accurate and efficient health management and monitoring are realized.

Description

Technical Field

[0001] This invention relates to the field of health management technology, specifically to a method and system for accurate health indicator prediction through multi-source data fusion. Background Technology

[0002] With the increasing demand for health management and personalized medicine, precision health monitoring has become an important research direction. Current technologies largely rely on single data sources or limited health indicators, making it difficult to comprehensively and dynamically reflect an individual's health status and its changing trends. To achieve accurate prediction and dynamic monitoring of individual health status, it is necessary to integrate multi-source data, covering multiple dimensions such as physiological, psychological, behavioral, and environmental factors, to comprehensively assess the user's health status.

[0003] Against this backdrop, this invention proposes a precise health indicator prediction method based on multi-source data fusion. By acquiring multidimensional user data and preprocessing it, a comprehensive dynamic health index is calculated by constructing bio-harmony features. Personalized health management suggestions are provided to users through threshold judgment and trend prediction. Furthermore, health trend prediction based on an LSTM model enables long-term monitoring and timely intervention of users' health status. Summary of the Invention

[0004] Based on the shortcomings of the prior art described above, the purpose of this invention is to provide a method and system for accurate health indicator prediction based on multi-source data fusion, so as to solve the above-mentioned technical problems.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for accurate health indicator prediction based on multi-source data fusion, comprising:

[0006] Acquire multidimensional data from users, including physiological, psychological, behavioral, and environmental data, and preprocess the multidimensional data.

[0007] Biological harmony features were constructed based on the preprocessed multidimensional data. These features include a comprehensive physiological health index, a psychological health fluctuation index, an activity intensity adaptation index, and an environmental stress index.

[0008] Calculate the comprehensive health dynamic index based on biological harmony characteristics;

[0009] The system assesses thresholds for the comprehensive health dynamic index and provides corresponding health recommendations.

[0010] Trend prediction is performed by constructing an LSTM model based on historical comprehensive health dynamic index;

[0011] Monitor your health status and health trend predictions in real time through a visual dashboard.

[0012] The present invention further specifies that the calculation logic for the comprehensive physiological health index is as follows: Among them, PHSI is the physiological health index, HR is heart rate, and BP is blood pressure. sys For systolic blood pressure, BP dia 1. Diastolic blood pressure; 2. SpO2. Blood oxygen saturation; 3. Weight; 4. Body temperature; 5. Normal body temperature; 6. k1; 7. k2; and 8. k3. Weighting coefficients.

[0013] The calculation logic for the mental health fluctuation index is as follows: Wherein, MHVI is the mental health fluctuation index, t is the time variable, S is the level of anxiety perception, E is the emotional stability, HRV is the heart rate variability, and α is the regulation parameter.

[0014] The calculation logic for the activity intensity adaptation index is as follows: Where AIAI is the activity intensity adaptation index, t0 is the total exercise duration, Act(t) is the activity intensity function, and HR is the activity intensity adaptation index. rest For resting heart rate, HR max With maximum heart rate, γ, τ, and η as adjustment parameters, the calculation logic of the activity intensity function Act(t) is as follows: Act(t) = W1·HR(t) + W2·a(t) + W3·MET(t), where HR(t) is the heart rate at the current time t, a(t) is the acceleration at the current time t, MET(t) is the metabolic equivalent at the current time t, and W1, W2, and W3 are weighting coefficients;

[0015] The calculation logic for the environmental stress index is as follows: Among them, ESI is the Environmental Stress Index, A PM2.5 The concentration of PM2.5 particulate matter in the air, H humidity For ambient humidity, T temperature For ambient temperature, N noise For noise level, T ideal θ represents the ideal temperature value, and θ is the adjustment parameter.

[0016] The present invention is further configured to calculate a comprehensive health dynamic index based on biological harmony characteristics, wherein the calculation logic of the comprehensive health dynamic index is as follows: Wherein, DHSI(t) is the Comprehensive Health Dynamic Index, PHSI is the Comprehensive Physiological Health Index, MHVI is the Mental Health Fluctuation Index, AIAI is the Activity Intensity Adaptation Index, ESI is the Environmental Stress Index, t is the time variable, λ is the decay factor, and δ... σ and σ are weighting coefficients.

[0017] The present invention is further configured to perform threshold judgment on the comprehensive health dynamic index. When the comprehensive health dynamic index is greater than the first threshold, it indicates that the current health status is good and it is recommended to maintain the existing health habits. When the comprehensive health dynamic index is greater than the second threshold and less than or equal to the first threshold, it indicates that the health status is average, and a first abnormal signal is generated, suggesting that the user adjust their lifestyle and have regular check-ups. When the comprehensive health dynamic index is less than or equal to the second threshold, it indicates that the health status is poor, and a second abnormal signal is generated, suggesting that emergency medical intervention and health tracking be carried out.

[0018] The present invention is further configured such that the dynamic adjustment steps of the first threshold and the second threshold include:

[0019] Collect basic user information, including age and medical history. The quantification logic for the severity of medical history is as follows: Where H represents the severity of the medical history, r represents the total number of diseases considered, and I represents the disease severity. q ) is an indicator of the existence of the qth disease, w q Let q be the weighting coefficient for the q-th disease;

[0020] The base threshold is adjusted based on the user's age and medical history. The calculation logic for the first threshold is as follows: τ1 is the first threshold, Let DHSI(t) be the first base threshold after normalization, Age be the user's age, μ be the age coefficient, and ω1 be the medical history sensitivity coefficient. The calculation logic for the second threshold is as follows: τ2 is the second threshold. ω1 is the second basic threshold after normalization of DHSI(t), and ω2 is the medical history sensitivity coefficient.

[0021] The present invention is further configured to dynamically optimize and adjust the first threshold based on the coupling of physiological rhythm and environment, the steps of which include:

[0022] Calculate the user's rhythm stability index (PRS);

[0023] Real-time analysis of the cohesion ratio (CR) between the environmental stress index (ESI) and heart rate variability (HRV);

[0024] The first threshold adjustment Δτ is generated by fusing PRS and CR.

[0025] The present invention is further configured such that the calculation logic of the rhythm stability index is as follows: PRS is the rhythm stability index, σ circadian The standard deviation of biological rhythms, μ circadian μ represents the mean of the biological rhythm. circadian The calculation logic is as follows: For physiological data at hour h, σ circadian The calculation logic is as follows:

[0026] When the ESI is greater than a preset threshold, coupling analysis is initiated to calculate the cohesion ratio between the ESI and HRV. The calculation logic for the cohesion ratio is as follows: CR stands for Coagulation Ratio;

[0027] The calculation logic for the first threshold adjustment Δτ is as follows: Δτ is the first threshold adjustment amount, T prs1 and T prs2 These are the threshold values ​​for the first and second rhythm stability indices, respectively, and T. cr1 The first cohesion ratio threshold, T cr2 The second threshold value of the absolute value of the cohesion ratio, T cr3 The third cohesion ratio threshold, w PRS For PRS weights, w ESI β1, β2, and β3 are the ESI weights, and w is the adjustment coefficient. PRS The calculation logic is as follows: δ1 and δ2 are adjustment coefficients, w ESI The calculation logic is as follows: w ESI =δ3·|CR|, where δ3 is the adjustment coefficient;

[0028] The calculation logic for optimizing and adjusting the first threshold is as follows: The first threshold is optimized and adjusted.

[0029] The present invention is further configured such that the step of constructing the LSTM model includes:

[0030] Obtain the user's historical comprehensive health dynamic index sequence {DHSI(tn),DHSI(t-n+1),...,DHSI(t)}, where n is the time window length, and standardize the sequence to obtain... The time window length n is determined based on the user data collection frequency;

[0031] Construct an LSTM network and train and optimize the LSTM model, adjusting the model parameters. The input layer is a standardized sequence of n time steps, and the output layer is the predicted value of the next m time steps. The prediction step size m is associated with the corresponding time of health intervention.

[0032] The latest n time steps Input a trained LSTM model and output the prediction result.

[0033] When the predicted value exceeds the preset safety threshold τ k times consecutively alert When the time comes, a pre-alarm is generated, the security threshold τ alert Dynamically adjusts based on user baseline;

[0034] Further analysis of the time intervals within s time periods Mutation gradient analysis is performed on the predicted value sequence to calculate the mutation energy integral within the time period [ts,t]. The calculation logic is as follows: Among them, E mut (t) is The mutation energy integral of the predicted value sequence within the time period [ts,t], where t is the current time, s is the backtracking time window length, and τ is the variable value. mut For time variables, for The gradient;

[0035] E mut (t) and the energy baseline E under the user's individual historical health status baseline Compare, when E is satisfied mut (t)>η e ·E baseline If an abnormal change occurs in the current health status, a confirmation alarm is generated, where η e This is the energy amplification factor;

[0036] A third abnormal signal is generated and a health intervention command is initiated only when both the preparatory alarm and the confirmed alarm are determined to have been triggered simultaneously.

[0037] The present invention is further configured such that the visual dashboard includes:

[0038] The real-time health status dashboard displays the dynamic changes of the comprehensive health dynamic index DHSI and its sub-indices, including the physiological health comprehensive index PHSI, the mental health fluctuation index MHVI, the activity intensity adaptation index AIAI, and the environmental stress index ESI.

[0039] The health trend prediction chart displays the prediction results of the LSTM model.

[0040] This invention also provides a multi-source data fusion-based accurate health indicator prediction system, the system comprising:

[0041] Data acquisition and processing module: used to acquire multidimensional data from users, including physiological data, psychological data, behavioral data and environmental data, and to preprocess the multidimensional data;

[0042] Feature construction module: used to construct biological harmony features based on preprocessed multidimensional data. Biological harmony features include physiological health comprehensive index, mental health fluctuation index, activity intensity adaptation index and environmental stress index.

[0043] Health status calculation module: used to calculate a comprehensive health dynamic index based on biological harmony characteristics;

[0044] Threshold judgment module: Used to judge the threshold of the comprehensive health dynamic index and provide corresponding health suggestions;

[0045] Trend prediction module: used to build an LSTM model based on historical comprehensive health dynamic index for trend prediction;

[0046] Visualization and Monitoring Module: Used to monitor health status and health trend prediction results in real time through visual dashboards.

[0047] This invention provides a method and system for accurate health indicator prediction based on multi-source data fusion. The method acquires multi-dimensional user data, including physiological, psychological, behavioral, and environmental data, and preprocesses this data. Based on the preprocessed multi-dimensional data, it constructs bio-harmony features, including a comprehensive physiological health index, a psychological health fluctuation index, an activity intensity adaptation index, and an environmental stress index. It then calculates a comprehensive dynamic health index based on these bio-harmony features. Threshold judgments are applied to the comprehensive dynamic health index, and corresponding health suggestions are provided. An LSTM model is constructed based on the historical comprehensive dynamic health index to predict trends. Finally, a visual dashboard monitors health status and health trend prediction results in real time. The beneficial effects include:

[0048] 1. Comprehensive Health Assessment: By integrating physiological, psychological, behavioral, and environmental data, it can comprehensively and systematically assess the user's health status. The comprehensive analysis of various data helps to identify potential health risks and provide more accurate health management solutions.

[0049] 2. Personalized health intervention: By constructing bioharmony characteristics and combining them with the user's historical health data, personalized health intervention plans can be realized. When the Comprehensive Health Dynamic Index (DHSI) reaches the threshold, the system automatically provides personalized health suggestions or intervention measures, thereby achieving more precise health management.

[0050] 3. Dynamic Threshold Adjustment: Taking into account the user's age, medical history, and health history, the threshold of health indicators can be dynamically adjusted to achieve flexible and personalized health status judgment. By optimizing and adjusting the false alarm rate, the accuracy of health monitoring can be further improved.

[0051] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0053] Figure 1 A flowchart illustrating a method for accurate health indicator prediction through multi-source data fusion, as shown in an exemplary embodiment of the present invention;

[0054] Figure 2 This is a schematic diagram illustrating the structure of a multi-source data fusion-based accurate health indicator prediction system, as an exemplary embodiment of the present invention. Detailed Implementation

[0055] The embodiments of the present invention will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention and not for limiting the scope of protection of the present invention.

[0056] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0057] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0058] Example 1

[0059] A method for accurate health indicator prediction through multi-source data fusion, such as... Figure 1 As shown, it includes:

[0060] Acquire multidimensional data from users, including physiological, psychological, behavioral, and environmental data, and preprocess the multidimensional data.

[0061] Biological harmony features were constructed based on the preprocessed multidimensional data. These features include a comprehensive physiological health index, a psychological health fluctuation index, an activity intensity adaptation index, and an environmental stress index.

[0062] Calculate the comprehensive health dynamic index based on biological harmony characteristics;

[0063] The system assesses thresholds for the comprehensive health dynamic index and provides corresponding health recommendations.

[0064] Trend prediction is performed by constructing an LSTM model based on historical comprehensive health dynamic index;

[0065] Monitor your health status and health trend prediction results in real time through a visual dashboard;

[0066] Specifically, physiological data includes heart rate, blood pressure saturation, and body temperature; psychological data includes mood fluctuations and stress levels; behavioral data includes activity intensity and duration; and environmental data includes air quality, temperature, humidity, and noise levels. Mood fluctuations can be obtained through changes in skin conductance and heart rate. Stress levels can be monitored through skin conductance responses, which reflect the activity of the autonomic nervous system, especially during stress or mood fluctuations, and can be measured using wearable devices. The acquired multidimensional data undergoes preprocessing, including data cleaning, standardization, denoising, and smoothing. Data cleaning removes incomplete or abnormal data using methods such as imputation, interpolation, deletion, and statistical methods. Data standardization converts data to a numerical range with the same dimensions using methods such as Z-score normalization and Min-Max normalization. Denoising and smoothing remove irrelevant noise, making the signal smoother and helping the model identify effective patterns. Filters are used to smooth data with large fluctuations, such as heart rate and blood pressure. Based on the preprocessed multidimensional data, health-related bioharmony features are calculated, including a comprehensive physiological health index and psychological... The system calculates a comprehensive dynamic health index (DHSI) by combining the Health Fluctuation Index, Activity Intensity Adaptation Index, and Environmental Stress Index with bioharmony characteristics, reflecting the user's overall health status in real time. The DHSI is normalized, and health thresholds are set based on these normalized values. Different thresholds represent different health states, and the system provides personalized health recommendations based on different DHSI value ranges. An LSTM model is constructed using historical DHSI values ​​to predict future health trends, identify potential health problems, and generate a third abnormality signal to provide early warnings. A visual dashboard displays the real-time trends of the DHSI and its sub-indices, including PHSI, MHVI, AIAI, and ESI, helping users understand their health status at any time. In addition to real-time monitoring, it also displays future health trend predictions based on the LSTM model, allowing users to see the trends in their health indices over a future period and take preventative and intervention measures in advance. Through the collection, preprocessing, feature extraction, dynamic index calculation, threshold judgment, trend prediction, and real-time monitoring of multidimensional data, the system provides users with a comprehensive and personalized health management tool.

[0067] The present invention further specifies that the calculation logic for the comprehensive physiological health index is as follows: Among them, PHSI is the physiological health index, HR is heart rate, and BP is blood pressure. sys For systolic blood pressure, BP dia The diastolic blood pressure, SpO2 (oxygen saturation), weight, T (body temperature), T0 (normal body temperature), and k1, k2, and k3 are weighting coefficients; the calculation logic of the mental health fluctuation index is as follows: Wherein, MHVI is the mental health fluctuation index, t is the time variable, S is the perceived level of anxiety, E is the emotional stability, HRV is the heart rate variability, and α is the regulation parameter; the calculation logic of the activity intensity adaptation index is as follows: Where AIAI is the activity intensity adaptation index, t0 is the total exercise duration, Act(t) is the activity intensity function, and HR is the activity intensity adaptation index. rest For resting heart rate, HR max With maximum heart rate, γ, τ, and η as adjustment parameters, the calculation logic of the activity intensity function Act(t) is: Act(t) = W1·HR(t) + W2·a(t) + W3·MET(t), where HR(t) is the heart rate at time t, a(t) is the acceleration at time t, MET(t) is the metabolic equivalent at time t, and W1, W2, and W3 are weighting coefficients; the calculation logic of the environmental stress index is: Among them, ESI is the Environmental Stress Index, A PM2.5 The concentration of PM2.5 particulate matter in the air, H humidity For ambient humidity, T temperature For ambient temperature, N noise For noise level, T idealHere, θ represents the ideal temperature value, and θ is the adjustment parameter. Specifically, the Physiological Health Index (PHSI) assesses an individual's physiological health status by integrating multiple physiological parameters; dynamic changes in PHSI can reflect changes in a person's health status in real time. The Mental Health Volatility Index (MHVI) is a dynamic indicator that measures fluctuations in an individual's mental health, combining factors such as stress level, emotional state, and heart rate variability. The Activity Intensity Adaptability Index (AIAI) measures an individual's adaptability to activity intensity over a period of time, primarily used to assess an individual's physiological adaptability under different intensity activity conditions. The Environmental Stress Index (ESI) is used to quantify and... Assess the health impact caused by environmental factors; the calculation logic for the perceived anxiety level S is: S = ε1·SCR(t) + σ1·SCL(t), where SCR(t) is the normalized peak value of the skin conductance response at the current time t, SCL(t) is the normalized long-term baseline average value of the skin conductance response, and ε and σ are weighting coefficients; the calculation logic for the emotional stability E is: E = ε2·ΔHR(t) + σ2·PeakSCR(t), where ΔHR(t) is the normalized heart rate change value, PeakSCR(t) is the normalized skin conductance response value, and ε1 and σ1 are weighting coefficients; activity intensity The function Act(t) is used to quantify the intensity of an individual's activity at time t; HR in the above physiological health comprehensive index calculation logic is the absolute value of the current heart rate minus the baseline heart rate; k1, k2, and k3 are used to adjust the influence of physiological data such as blood oxygen saturation, heart rate, blood pressure, and body temperature, and their specific values ​​depend on the experimental data; ε1 and σ1 are used to adjust the contribution of the peak and average values ​​of the skin conductance response to the degree of anxiety perception, respectively, with a value range of [0,1]; ε2 and σ2 are used to adjust the influence of heart rate and skin conductance response on emotional stability, respectively, with a value range of [0,1]; α is used to adjust the degree of anxiety perception. The relationship between the degree and the mental health fluctuation index is defined as follows: γ is used to regulate the effect of activity intensity on the growth of the index, with a value range of [1,3]; τ is used to control the adaptive effect of time on activity intensity, with a value range of [0,10]; η is used to control the change of the activity intensity function over time, with a value range of [0,3]; W1, W2, and W3 are used to control the contribution of heart rate, acceleration, and metabolic equivalent to activity intensity, respectively, with a value range of [0,1], and the sum of W1, W2, and W3 is 1; θ is used to control the influence of environmental humidity on environmental stress, with a value range of [0.1,1].

[0068] The present invention is further configured to calculate a comprehensive health dynamic index based on biological harmony characteristics, wherein the calculation logic of the comprehensive health dynamic index is as follows: Wherein, DHSI(t) is the Comprehensive Health Dynamic Index, PHSI is the Comprehensive Physiological Health Index, MHVI is the Mental Health Fluctuation Index, AIAI is the Activity Intensity Adaptation Index, ESI is the Environmental Stress Index, t is the time variable, λ is the decay factor, and δ... σ and σ are weighting coefficients; specifically, the Comprehensive Health Dynamic Index (DHSI)(t) is used to assess an individual's health status in multiple dimensions, including four dimensions: physical, psychological, activity, and environment. DHSI(t) is calculated after standardizing PHSI, MHVI, AIAI, and ESI. When the DHSI(t) value is large, it indicates a better health status, reflecting that physical health, mental health, activity adaptation, and environmental factors are at a relatively ideal level; λ is used to control the decline of health status over time, with a value range of [0,1]; δ is used to adjust the contribution of the Physical Health Comprehensive Index (PHSI) to the Comprehensive Health Dynamic Index (DHSI)(t), with a value range of [0.5,1]. The MHVI (Mental Health Volatility Index) is used to adjust the contribution of the DHSI (Health Dynamic Index) to the comprehensive health dynamic index, with a value range of [0.3, 0.7]. σ is used to adjust the contribution of the AIAI (Activity Intensity Adaptation Index) and ESI (Environmental Stress Index) to the DHSI (Health Dynamic Index), with a value range of [0.1, 0.5]. The DHSI (Health Dynamic Index) can comprehensively reflect an individual's physiological, psychological, activity, and environmental health status, providing a comprehensive assessment of an individual's health status.

[0069] The invention is further configured to perform threshold judgment on the comprehensive health dynamic index. When the comprehensive health dynamic index is greater than a first threshold, it indicates that the current health status is good, and it is recommended to maintain the existing health habits. When the comprehensive health dynamic index is greater than a second threshold and less than or equal to the first threshold, it indicates that the health status is average, generating a first abnormal signal, and recommending that the user adjust their lifestyle and have regular check-ups. When the comprehensive health dynamic index is less than or equal to the second threshold, it indicates that the health status is poor, generating a second abnormal signal, and recommending emergency medical intervention and health tracking. Specifically, appropriate health advice is provided to users based on changes in the DHSI(t) value. According to different values ​​of the comprehensive health dynamic index, it is possible to assess health status, generate health signals, and provide intervention mechanisms. The first threshold is set as a relatively high level of health. The system has two thresholds. A first threshold indicates good health, meaning the user doesn't need to drastically change their lifestyle and is advised to maintain existing healthy habits, including a healthy diet, exercise, and lifestyle. A second threshold represents a moderate health level. When the DHSI(t) value is greater than the second threshold but less than or equal to the first threshold, a first abnormal signal is generated, indicating that the user's health is generally poor, possibly due to fatigue or a sub-healthy state. Adjustments to lifestyle habits and regular health checkups are recommended. A second abnormal signal is generated when the DHSI(t) value is less than or equal to the second threshold, indicating poor health and requiring emergency medical intervention and continuous monitoring of health changes. Personalized health advice and interventions are provided based on different health states to help users improve their lifestyle habits and enhance their health.

[0070] The present invention is further configured such that the dynamic adjustment steps of the first threshold and the second threshold include:

[0071] Collect basic user information, including age and medical history. The quantification logic for the severity of medical history is as follows: Where H represents the severity of the medical history, r represents the total number of diseases considered, and I represents the disease severity. q ) is an indicator of the existence of the qth disease, w q Let q be the weighting coefficient for the q-th disease;

[0072] The base threshold is adjusted based on the user's age and medical history. The calculation logic for the first threshold is as follows: τ1 is the first threshold, Let DHSI(t) be the first base threshold after normalization, Age be the user's age, μ be the age coefficient, and ω1 be the medical history sensitivity coefficient. The calculation logic for the second threshold is as follows: τ2 is the second threshold. ω2 is the second basic threshold after normalization of DHSI(t), and ω2 is the medical history sensitivity coefficient. Specifically, the severity of medical history H is used to quantify the user's medical history information, comprehensively considering different types of diseases and their severity, so as to provide a reference for the individual's health status in health monitoring. The larger the value of H, the greater the impact of historical diseases on the current health status. The first threshold τ1 and the first threshold τ2 used for health status discrimination are adjusted in combination with the user's age and medical history, so as to more personally assess the user's health status and provide highly adaptive health management suggestions. r usually represents the total number of major diseases that are of concern during individual health monitoring; I(Disease) q Used to indicate the presence of a disease or the intensity of its effects; q The values ​​are used to quantify and measure the impact of each disease on an individual's health. The values ​​depend on the nature of the disease, the difficulty of treatment, and the actual impact on health. The specific values ​​are determined based on the advice of medical experts, clinical data, or historical experience of medical records. μ is used to quantify the impact of age on the health threshold, with a value range of [0, 0.01]. ω1 and ω2 are used to determine the degree of influence of medical history on health assessment, with a value range of [0, 0.1]. Adjusting the threshold by medical history and age helps to identify high-risk groups and conduct earlier interventions to reduce the occurrence of health risks.

[0073] The present invention is further configured to dynamically optimize and adjust the first threshold based on the coupling of physiological rhythm and environment, the steps of which include:

[0074] The user's rhythm stability index (PRS) is calculated; the cohesion ratio (CR) between the environmental stress index (ESI) and heart rate variability (HRV) is analyzed in real time; a first threshold adjustment Δτ is generated by fusing PRS and CR; the present invention is further configured such that the calculation logic of the rhythm stability index is as follows: PRS is the rhythm stability index, σ circadian The standard deviation of biological rhythms, μ circadian μ represents the mean of the biological rhythm. circadian The calculation logic is as follows: For physiological data at hour h, σ circadian The calculation logic is as follows: When the ESI is greater than a preset threshold, coupling analysis is initiated to calculate the cohesion ratio between the ESI and HRV. The calculation logic for the cohesion ratio is as follows: CR is the cohesion ratio; the calculation logic for the first threshold adjustment Δτ is as follows: Δτ is the first threshold adjustment amount, T prs1 and T prs2 These are the threshold values ​​for the first and second rhythm stability indices, respectively, and T. cr1 The first cohesion ratio threshold, T cr2The second threshold value of the absolute value of the cohesion ratio, T cr3 The third cohesion ratio threshold, w PRS For PRS weights, w ESI β1, β2, and β3 are the ESI weights, and w is the adjustment coefficient. PRS The calculation logic is as follows: δ1 and δ2 are adjustment coefficients, w ESI The calculation logic is as follows: w ESI =δ3·|CR|, where δ3 is the adjustment coefficient; the calculation logic for the optimized adjustment of the first threshold is as follows: The first threshold has been optimized and adjusted; specifically, the Rhythm Stability Index (PRS) is used to measure the stability of an individual's biological rhythms. A low PRS value may indicate a disruption in the user's biological rhythms, which could subsequently affect their health; the standard deviation of the biological rhythm σ... circadian μ is used to measure the degree of dispersion of data from the mean. circadian The average level of physiological data over a 24-hour period is used, where the physiological data is heart rate or body temperature; the cohesion ratio (CR) represents the coupling strength between ESI and HRV. A high CR value may indicate that environmental stress significantly affects heart rate changes, thus reflecting the body's stress response; the first threshold adjustment Δτ is calculated based on the rhythm stability index (PRS) and the cohesion ratio (CR). The first threshold is adjusted because it is designed to compare more sensitive health data, and relaxing the threshold reduces unnecessary alarms; T prs1 Setting it to a low value indicates that health intervention is needed when PRS falls below that value; T prs2 Set to a relatively high value to determine whether an individual's circadian rhythm is relatively stable; T cr1 The CR value is used to determine if the heart rate variability is too low; a value below this indicates that environmental stress has a relatively small impact on heart rate variability. cr2 Used to further adjust the effects of CR and ESI; T cr3The threshold is determined by β1, β2, and β3, which control the magnitude of threshold changes under different conditions, with values ​​ranging from [0,1]. δ1 controls the sensitivity of PRS changes, with values ​​ranging from [0.1,5]. δ2 controls the inflection point of the PRS curve, with values ​​ranging from [0,1]. δ3 adjusts the influence of ESI on threshold adjustment, with values ​​ranging from [0,0.5]. The threshold is dynamically adjusted based on the values ​​of the rhythm stability index PRS and the cohesion ratio CR, allowing the system to make real-time adjustments based on real-time health data. It should be noted that when ESI is less than or equal to the preset threshold, the coupling analysis adjustment is not satisfied, and it is not necessary to calculate the first threshold adjustment amount Δτ. The threshold judgment of the comprehensive health dynamic index DHSI(t) is performed according to the original first threshold τ1.

[0075] The present invention is further configured such that the step of constructing the LSTM model includes:

[0076] Obtain the user's historical comprehensive health dynamic index sequence {DHSI(tn),DHSI(t-n+1),...,DHSI(t)}, where n is the time window length, and standardize the sequence to obtain... The time window length *n* is determined based on the user data collection frequency. Specifically, the time window length *n* determines the length of the data sequence input to the LSTM model. If data is collected once a day, then *n* = 7, indicating that the data input to the LSTM model is from the past 7 days. If data is collected once an hour, then *n* = 24, indicating that the data input to the LSTM model is from the past 24 hours. The data is standardized using Z-score normalization. The data here refers to the comprehensive health dynamic index data:

[0077] An LSTM network is constructed and trained and optimized, adjusting the model parameters. The input layer is a standardized sequence of n time steps, and the output layer is the predicted value for the next m time steps. The prediction step size m is associated with the corresponding time of the health intervention. Specifically, the input layer receives data from the past n time steps, i.e., the past n days or n hours. The sequence has an input data dimension of n, where each time step represents a sequence of data. The goal of the output layer is to predict the comprehensive health dynamic index over the next m time steps, where m represents the prediction step size. This is usually associated with the timing of health interventions. For example, if we want to predict the health status over the next 7 days, then m = 7. The model is trained to adjust its internal parameters so that it can accurately predict the future health status. The training process usually uses historical data to minimize the error between the predicted and the actual values. This is an existing technique and will not be elaborated on here.

[0078] The latest n time steps Input a trained LSTM model and output the prediction result. Specifically, given the comprehensive health dynamic index data for the past 7 days, the LSTM model will output the comprehensive health dynamic index for the next 7 days. That is, the prediction of the comprehensive health dynamic index for the next time step;

[0079] When the predicted value exceeds the preset safety threshold τ k times consecutively alert When the time comes, a pre-alarm is generated, the security threshold τ alert The system dynamically adjusts based on the user's baseline; specifically, when the model's predicted comprehensive health dynamic index exceeds the preset τ for k consecutive times... alert The threshold will automatically generate a preliminary alert, indicating a trend of deteriorating health status; the safety threshold τ alert The steps for dynamically adjusting based on the user's baseline include: obtaining the user's historical Health Dynamic Index (DHSI) data over a past period, calculating the mean and standard deviation of the historical DHSI data, and then setting a safety threshold τ. alert Safety threshold τ alert The calculation logic is as follows: τ alert =μ baseline +η a ·σ baseline μ baseline The mean and σ of the historical DHSI data baseline The standard deviation and η of the historical DHSI data a η is the adjustment coefficient; a Used to control the tolerance of the safety threshold to individual variability, with a value range of [1,2];

[0080] Further analysis of the time intervals within s time periods Mutation gradient analysis is performed on the predicted value sequence to calculate the mutation energy integral within the time period [ts,t]. The calculation logic is as follows: Among them, E mut (t) is The mutation energy integral of the predicted value sequence within the time period [ts,t], where t is the current time, s is the backtracking time window length, and τ is the variable value. mut For time variables, for The gradient; specifically, to verify whether a real mutation has occurred in the health status, after generating the preliminary alarm, further mutation gradient analysis is performed within the most recent s prediction times, and the mutation energy integral is calculated;

[0081] E mut (t) and the energy baseline E under the user's individual historical health status baseline Compare, when E is satisfied mut (t)>ηe ·E baseline If an abnormal change occurs in the current health status, a confirmation alarm is generated, where η e This represents the energy amplification factor; specifically, when the above criteria are met, it indicates that the predicted sequence has recently experienced a drastic, non-stationary fluctuation in its health status, generating a confirmation alarm and improving the reliability of anomaly identification; Energy benchmark E baseline It can be obtained by analyzing the fluctuation rate of the comprehensive health dynamic index under the user's historical stable health status, calculated using methods such as sliding window mean square error integral, or by combining medical assessment data for parameter calibration, to reflect the normal fluctuation energy level of an individual in a healthy state; η e Used to determine the current mutation energy E mut (t) Is it significantly higher than the user's historical average level E? baseline The value range is [1.2, 2].

[0082] A third abnormal signal is generated and a health intervention instruction is initiated only when both the pre-alarm and the confirmed alarm are triggered simultaneously. Specifically, to avoid false alarms and over-intervention, an abnormal health status is considered to be effectively confirmed only when both of the above conditions are met simultaneously, thereby generating a third abnormal signal and initiating a health intervention instruction. The intervention instruction includes pushing personalized health advice to the user and activating third-party services.

[0083] The present invention is further configured such that the visual dashboard includes:

[0084] The real-time health status dashboard displays the dynamic changes of the Comprehensive Health Dynamic Index (DHSI) and its sub-indices, including the Physiological Health Index (PHSI), Mental Health Fluctuation Index (MHVI), Activity Intensity Adaptation Index (AIAI), and Environmental Stress Index (ESI). A health trend prediction chart displays the prediction results of the LSTM model. Specifically, the real-time health status dashboard displays the user's health status in real time and dynamically shows the changing trends of the Comprehensive Health Dynamic Index and its sub-indices. Through real-time monitoring and display, users can understand their health status in a timely manner and make appropriate adjustments based on changes. The health trend prediction chart, by displaying health prediction results based on the LSTM model, helps users understand health trends over a future period. The visual dashboard provides personalized health suggestions and interventions based on individual health data and prediction results, thereby improving the efficiency and accuracy of health management.

[0085] Example 2

[0086] Please see Figure 2 This exemplary multi-source data fusion-based accurate health indicator prediction system includes:

[0087] Data acquisition and processing module: used to acquire multidimensional data from users, including physiological data, psychological data, behavioral data and environmental data, and to preprocess the multidimensional data;

[0088] Feature construction module: used to construct biological harmony features based on preprocessed multidimensional data. Biological harmony features include physiological health comprehensive index, mental health fluctuation index, activity intensity adaptation index and environmental stress index.

[0089] Health status calculation module: used to calculate a comprehensive health dynamic index based on biological harmony characteristics;

[0090] Threshold judgment module: Used to judge the threshold of the comprehensive health dynamic index and provide corresponding health suggestions;

[0091] Trend prediction module: used to build an LSTM model based on historical comprehensive health dynamic index for trend prediction;

[0092] Visualization and Monitoring Module: Used to monitor health status and health trend prediction results in real time through visual dashboards.

[0093] It should be noted that the multi-source data fusion-based accurate health indicator prediction system and the multi-source data fusion-based accurate health indicator prediction method provided in the above embodiments belong to the same concept. The specific operation methods of each module and unit have been described in detail in the method embodiments and will not be repeated here. In practical applications, the multi-source data fusion-based accurate health indicator prediction system provided in the above embodiments can be assigned to different functional modules as needed, that is, the internal structure of the system can be divided into different functional modules to complete all or part of the functions described above. This is not a limitation here.

[0094] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0095] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0096] In this application, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or multiple items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0097] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0098] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0099] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0100] In the several embodiments provided in this application, it should be understood that the disclosed system can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0101] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0102] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0103] If the aforementioned functions are implemented as 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 this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0104] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for accurate health indicator prediction through multi-source data fusion, characterized in that, include: Acquire multidimensional data from users, including physiological, psychological, behavioral, and environmental data, and preprocess the multidimensional data. Biological harmony features were constructed based on the preprocessed multidimensional data. These features include a comprehensive physiological health index, a psychological health fluctuation index, an activity intensity adaptation index, and an environmental stress index. Calculate the comprehensive health dynamic index based on biological harmony characteristics; The system assesses thresholds for the comprehensive health dynamic index and provides corresponding health recommendations. Trend prediction is performed by constructing an LSTM model based on historical comprehensive health dynamic index; Monitor your health status and health trend predictions in real time through a visual dashboard.

2. The method for accurate health indicator prediction based on multi-source data fusion according to claim 1, characterized in that, The calculation logic for the comprehensive physiological health index is as follows: Among them, PHSI is the physiological health index, HR is heart rate, and BP is blood pressure. sys For systolic blood pressure, BP dia 1. Diastolic blood pressure; 2. SpO2. Blood oxygen saturation; 3. Weight; 4. Body temperature; 5. Normal body temperature; 6. k1; 7. k2; and 8. k3. Weighting coefficients. The calculation logic for the mental health fluctuation index is as follows: Wherein, MHVI is the mental health fluctuation index, t is the time variable, S is the level of anxiety perception, E is the emotional stability, HRV is the heart rate variability, and α is the regulation parameter. The calculation logic for the activity intensity adaptation index is as follows: Where AIAI is the activity intensity adaptation index, t0 is the total exercise duration, Act(t) is the activity intensity function, and HR is the activity intensity adaptation index. rest For resting heart rate, HR max With maximum heart rate, γ, τ, and η as adjustment parameters, the calculation logic of the activity intensity function Act(t) is as follows: Act(t) = W1·HR(t) + W2·a(t) + W3·MET(t), where HR(t) is the heart rate at the current time t, a(t) is the acceleration at the current time t, MET(t) is the metabolic equivalent at the current time t, and W1, W2, and W3 are weighting coefficients; The calculation logic for the environmental stress index is as follows: Among them, ESI is the Environmental Stress Index, A PM2.5 The concentration of PM2.5 particulate matter in the air, H humidity For ambient humidity, T temperature For ambient temperature, N noise For noise level, T ideal θ represents the ideal temperature value, and θ is the adjustment parameter.

3. The method for accurate health indicator prediction based on multi-source data fusion according to claim 1, characterized in that, The comprehensive health dynamic index is calculated based on biological harmony characteristics. The calculation logic of the comprehensive health dynamic index is as follows: Wherein, DHSI(t) is the Comprehensive Health Dynamic Index, PHSI is the Comprehensive Physiological Health Index, MHVI is the Mental Health Fluctuation Index, AIAI is the Activity Intensity Adaptation Index, ESI is the Environmental Stress Index, t is the time variable, λ is the decay factor, and δ... σ and σ are weighting coefficients.

4. The method for accurate health indicator prediction based on multi-source data fusion according to claim 1, characterized in that, A threshold judgment is made on the comprehensive health dynamic index. When the comprehensive health dynamic index is greater than the first threshold, it indicates that the current health status is good and it is recommended to maintain the existing health habits. When the comprehensive health dynamic index is greater than the second threshold and less than or equal to the first threshold, it indicates that the health status is generally good, generating the first abnormal signal. It is recommended that the user adjust their lifestyle and have regular check-ups. When the comprehensive health dynamic index is less than or equal to the second threshold, it indicates poor health status, generates a second abnormal signal, and recommends emergency medical intervention and health tracking.

5. The method for accurate health indicator prediction based on multi-source data fusion according to claim 4, characterized in that, The dynamic adjustment steps for the first and second thresholds include: Collect basic user information, including age and medical history. The quantification logic for the severity of medical history is as follows: Where H represents the severity of the medical history, r represents the total number of diseases considered, and I represents the disease severity. q ) is an indicator of the existence of the qth disease, w q Let q be the weighting coefficient for the q-th disease; The base threshold is adjusted based on the user's age and medical history. The calculation logic for the first threshold is as follows: τ1 is the first threshold, Let DHSI(t) be the first base threshold after normalization, Age be the user's age, μ be the age coefficient, and ω1 be the medical history sensitivity coefficient. The calculation logic for the second threshold is as follows: τ2 is the second threshold. ω1 is the second basic threshold after normalization of DHSI(t), and ω2 is the medical history sensitivity coefficient.

6. The method for accurate health indicator prediction based on multi-source data fusion according to claim 5, characterized in that, The first threshold is dynamically optimized and adjusted based on the coupling of physiological rhythms and the environment. The steps include: Calculate the user's rhythm stability index (PRS); Real-time analysis of the cohesion ratio (CR) between the environmental stress index (ESI) and heart rate variability (HRV); The first threshold adjustment Δτ is generated by fusing PRS and CR.

7. The method for accurate health indicator prediction based on multi-source data fusion according to claim 6, characterized in that, The calculation logic for the rhythm stability index is as follows: PRS is the rhythm stability index, σ circadian The standard deviation of biological rhythms, μ circadian μ represents the mean of the biological rhythm. circadian The calculation logic is as follows: For physiological data at hour h, σ circadian The calculation logic is as follows: When the ESI is greater than a preset threshold, coupling analysis is initiated to calculate the cohesion ratio between the ESI and HRV. The calculation logic for the cohesion ratio is as follows: CR stands for Coagulation Ratio; The calculation logic for the first threshold adjustment Δτ is: Δτ = Δτ is the first threshold adjustment amount, T prs1 and T prs2 These are the threshold values ​​for the first and second rhythm stability indices, respectively, and T. cr1 The first cohesion ratio threshold, T cr2 The second threshold value of the absolute value of the cohesion ratio, T cr3 The third cohesion ratio threshold, w PRS For PRS weights, w ESI β1, β2, and β3 are the ESI weights, and w is the adjustment coefficient. PRS The calculation logic is as follows: δ1 and δ2 are adjustment coefficients, w ESI The calculation logic is as follows: w ESI =δ3·|CR|, where δ3 is the adjustment coefficient; The calculation logic for optimizing and adjusting the first threshold is as follows: The first threshold is optimized and adjusted.

8. The method for accurate health indicator prediction based on multi-source data fusion according to claim 1, characterized in that, The steps for constructing the LSTM model include: Obtain the user's historical comprehensive health dynamic index sequence {DHSI(tn),DHSI(t-n+1),...,DHSI(t)}, where n is the time window length, and standardize the sequence to obtain... The time window length n is determined based on the user data collection frequency; Construct an LSTM network and train and optimize the LSTM model, adjusting the model parameters. The input layer is a standardized sequence of n time steps, and the output layer is the predicted value of the next m time steps. The prediction step size m is associated with the corresponding time of health intervention. The latest n time steps Input a trained LSTM model and output the prediction result. When the predicted value exceeds the preset safety threshold τ k times consecutively alert When the time comes, a pre-alarm is generated, the security threshold τ alert Dynamically adjusts based on user baseline; Further analysis of the time intervals within s time periods Mutation gradient analysis is performed on the predicted value sequence to calculate the mutation energy integral within the time period [ts,t]. The calculation logic is as follows: Among them, E mut (t) is The mutation energy integral of the predicted value sequence within the time period [ts,t], where t is the current time, s is the backtracking time window length, and τ is the variable value. mut For time variables, for The gradient; E mut (t) and the energy baseline E under the user's individual historical health status baseline Compare, when E is satisfied mut (t)>η e ·E baseline If an abnormal change occurs in the current health status, a confirmation alarm is generated, where η e This is the energy amplification factor; A third abnormal signal is generated and a health intervention command is initiated only when both the preparatory alarm and the confirmed alarm are determined to have been triggered simultaneously.

9. The method for accurate health indicator prediction based on multi-source data fusion according to claim 1, characterized in that, The visualization dashboard includes: The real-time health status dashboard displays the dynamic changes of the comprehensive health dynamic index DHSI and its sub-indices, including the physiological health comprehensive index PHSI, the mental health fluctuation index MHVI, the activity intensity adaptation index AIAI, and the environmental stress index ESI. The health trend prediction chart displays the prediction results of the LSTM model.

10. A multi-source data fusion-based accurate health indicator prediction system, used to implement the multi-source data fusion-based accurate health indicator prediction method according to any one of claims 1-9, characterized in that, include: Data acquisition and processing module: used to acquire multidimensional data from users, including physiological data, psychological data, behavioral data and environmental data, and to preprocess the multidimensional data; Feature construction module: used to construct biological harmony features based on preprocessed multidimensional data. Biological harmony features include physiological health comprehensive index, mental health fluctuation index, activity intensity adaptation index and environmental stress index. Health status calculation module: used to calculate a comprehensive health dynamic index based on biological harmony characteristics; Threshold judgment module: Used to judge the threshold of the comprehensive health dynamic index and provide corresponding health suggestions; Trend prediction module: used to build an LSTM model based on historical comprehensive health dynamic index for trend prediction; Visualization and Monitoring Module: Used to monitor health status and health trend prediction results in real time through visual dashboards.

Images