Mental health early warning method and system based on intelligence

Abstract

The invention relates to the technical field of intelligent psychological health monitoring and early warning, and particularly discloses a psychological health early warning method and system based on intelligence, and the method comprises the steps: collecting the skin temperature fluctuation and body movement frequency data of a user in real time through a wearable device in a psychological health monitoring period; independently reflecting autonomic nerve activity intensity and behavior rhythm coordination under emotion induction; key features are extracted in combination with Haar wavelet transform and fast Fourier transform and fused and input into the trained gradient boosting tree model, and quantitative evaluation and intelligent judgment of the psychological stability state of the user are achieved; whether a grading early warning mechanism is triggered or not is judged according to the evaluation result, first-level or second-level mental health early warning is provided, and the system comprises a data acquisition module, an autonomic nervous system response evaluation module, a psychological coordination evaluation module, a psychological state stability evaluation module and an early warning module and has the advantages of being non-invasive, high in precision and high in real-time performance.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent mental health monitoring and early warning, and in particular to an intelligent-based mental health early warning method and system. Background Art

[0002] With the accelerated pace of modern society and the increasing pressures of life, mental health issues are becoming increasingly prominent and have become a significant factor affecting public health. Traditional mental health assessment methods primarily rely on subjective questionnaires, clinical interviews, or professional psychological tests. These methods have limitations such as long assessment cycles, reliance on individual subjective expressions, and difficulty in real-time monitoring. They struggle to meet the needs of individualized, dynamic mental state tracking and early warning. In recent years, with the development of wearable devices and physiological signal acquisition technology, mental state recognition based on physiological indicators has gradually become a research hotspot. Previous studies have shown that skin temperature fluctuations are closely related to autonomic nervous system activity and can reflect individual emotional changes. Body movement frequency, on the other hand, reflects the stability of behavioral rhythms and is an important external manifestation of psychological coordination.

[0003] The existing technology has the following deficiencies:

[0004] Current mental health early warning systems generally rely on common physiological or behavioral characteristics such as heart rate, galvanic skin response, and voice intonation to identify mental states, neglecting potentially diagnostically valuable atypical physiological parameters such as asymmetric changes in local skin temperature and the duration of micromovements. Existing technologies lack the ability to synchronously acquire these signals and multimodal fusion analysis mechanisms, resulting in a high rate of missed detections in the early identification of psychological stress or emotional fluctuations. Furthermore, most systems fail to account for the temporal dynamics of individual physiological responses, making it difficult to respond promptly and accurately to sudden emotional fluctuations, thus compromising the practicality and robustness of early warning systems. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent mental health early warning method and system to solve the problems in the above background.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] The intelligent-based mental health early warning method includes the following steps:

[0008] S1: During the mental health monitoring period, the user's skin temperature data and body movement frequency data are collected in real time;

[0009] S2: Obtain the user's skin temperature fluctuation data and calculate the differential temperature characteristic value based on the abnormal degree of skin differential temperature fluctuation to evaluate the intensity of the autonomic nervous system response under emotion induction;

[0010] S3: Analyze the rhythm stability of body movement frequency, calculate the characteristic value of body movement rhythm deviation, and evaluate the user's psychological coordination;

[0011] S4: Perform fusion analysis on the differential temperature characteristic value and the body movement rhythm deviation characteristic value, and judge whether the current psychological state is at an unstable level based on the analysis results;

[0012] S5: If it is determined to be a psychologically unstable state, the system will automatically activate the early warning mechanism.

[0013] As a further embodiment of the present invention, the assessment of the intensity of the autonomic nervous system response under emotion induction specifically includes:

[0014] During the mental health monitoring period, the user's skin temperature fluctuation data is obtained, and the differential temperature characteristic value is calculated based on the abnormal degree of skin differential temperature fluctuation. It is determined whether the differential temperature characteristic value is greater than or equal to the preset threshold. If so, the autonomic nervous system response under emotion induction is abnormal. If not, the autonomic nervous system response under emotion induction is normal.

[0015] As a further solution of the present invention: the process of obtaining the differential temperature characteristic value is:

[0016] During the mental health monitoring period, the user's skin temperature fluctuation data is collected to form the original temperature time series;

[0017] Calculate the absolute value of the temperature difference between adjacent original temperature times, and integrate the obtained absolute value of the temperature difference into a differential temperature data sequence;

[0018] Perform a Haar wavelet transform on the differential temperature data sequence to obtain a detail coefficient sequence and calculate the energy value of the detail coefficient;

[0019] The average energy value of all detail coefficients is calculated by using the mean calculation expression, and the ratio of the average energy value of all detail coefficients to the total energy value of all detail coefficients is calculated to obtain the differential temperature characteristic value.

[0020] As a further solution of the present invention: the evaluation of the user's psychological coordination specifically includes:

[0021] During the mental health monitoring period, the user's body movement frequency data is obtained, and the body movement rhythm deviation characteristic value is calculated based on the degree of deviation of the user's body movement frequency. It is determined whether the body movement rhythm deviation characteristic value is greater than or equal to the preset threshold. If so, the user's psychological coordination is abnormal; if not, the user's psychological coordination is normal.

[0022] As a further solution of the present invention: the process of obtaining the body movement rhythm deviation characteristic value is:

[0023] During the mental health monitoring period, the user's body movement frequency data is collected to form an original body movement time series;

[0024] Perform fast Fourier transform on the body motion frequency time series to obtain the frequency domain signal, and extract the main frequency components in the frequency domain signal;

[0025] The absolute value of the difference between the main frequency component and the preset expected frequency is calculated and the ratio is calculated with the maximum frequency component to obtain the body movement rhythm deviation characteristic value.

[0026] As a further solution of the present invention: the fusion analysis of the differential temperature characteristic value and the body movement rhythm deviation characteristic value specifically includes:

[0027] During the mental health monitoring cycle, the user's psychological differential temperature characteristic values and body movement rhythm deviation characteristic values are obtained, and the differential temperature characteristic values and body movement rhythm deviation characteristic values are constructed into a comprehensive characteristic vector as the input of the mental health prediction model. The mental health prediction model is trained, and based on the trained mental health prediction model, the user's psychological stability score is output. Based on the user's psychological stability score, whether the current user's psychological state is stable is evaluated. The mental health prediction model is a gradient boosting tree model.

[0028] As a further solution of the present invention: the training process of the mental health prediction model is:

[0029] The comprehensive feature vector is obtained and input into the gradient boosting tree model to minimize the error between the predicted user psychological stability score and the actual user psychological stability score. This is used as the training target of the gradient boosting tree model. A cross-validation strategy is adopted, combined with grid search to optimize the hyperparameter configuration to improve the generalization ability and prediction accuracy of the model. The gradient descent method is used to minimize the loss function, and the decision tree set is iteratively updated to gradually enhance the model's ability to distinguish between psychologically stable and unstable states, thereby outputting an accurate user psychological stability score.

[0030] As a further solution of the present invention: the determining whether the current mental state is at an unstable level specifically includes:

[0031] During the mental health monitoring period, it is determined whether the psychological stability score of the current mental state is greater than or equal to a preset threshold. If so, the current user's mental state is at a stable level; if not, the current user's mental state is at an unstable level.

[0032] As a further solution of the present invention: if the patient is determined to be in a psychologically unstable state, the system automatically activates an early warning mechanism, specifically including:

[0033] According to the unstable level of the current user's mental state, the number of cycles of the user's mental state instability is obtained. If the number of cycles of the user's mental state instability is less than two monitoring cycles, the user's mental state is determined to be a first-level abnormal state, and the current user is marked and a mental health warning is issued. If the number of cycles of the user's mental state instability is greater than or equal to two monitoring cycles, the user's mental state is determined to be a second-level abnormal state, and a mental health check is immediately performed on the current user, and a serious warning is issued.

[0034] An intelligent mental health early warning system, including:

[0035] A data acquisition module, which collects the user's skin temperature data and body movement frequency data in real time during the mental health monitoring period;

[0036] a user autonomic nervous system response evaluation module, which obtains the user's skin temperature fluctuation data and calculates a differential temperature characteristic value based on the abnormal degree of skin differential temperature fluctuation, for use in evaluating the intensity of the autonomic nervous system response under emotion induction;

[0037] A user psychological coordination evaluation module, which evaluates the user's psychological coordination by analyzing the rhythm stability of body movement frequency and calculating the body movement rhythm deviation characteristic value;

[0038] A user psychological state stability assessment module, which performs a fusion analysis on the differential temperature characteristic value and the body movement rhythm deviation characteristic value, and determines whether the current psychological state is at an unstable level based on the analysis result;

[0039] The early warning module automatically activates the early warning mechanism if the early warning module determines that the patient is in a state of psychological instability.

[0040] Beneficial effects of the present invention:

[0041] (1) The present invention constructs a non-invasive mental health assessment framework based on skin temperature fluctuations and body movement frequency through the fusion analysis of multimodal physiological signals. Specifically, the system collects the user's skin temperature data and three-dimensional body movement frequency signals in real time during the continuous monitoring period, reflecting the dynamic response of the autonomic nervous system and the stability of behavioral rhythms under emotion induction respectively; further, the Haar wavelet transform is used to perform time-frequency decomposition of the differential temperature series, extracting high-frequency energy features to characterize the instantaneous fluctuation intensity during the body temperature regulation process; at the same time, the fast Fourier transform is used to perform frequency domain modeling on the body movement frequency time series, identifying the main movement rhythm and calculating the degree of deviation from the ideal frequency pattern to form the body movement rhythm deviation feature. Finally, the two types of features are fused and input into the mental health prediction model constructed based on the gradient boosting tree, and the intelligent discrimination of the mental stability state is achieved through the supervised learning mechanism. Compared with traditional psychological assessment methods that rely on subjective reports or clinical observations, this invention can capture subtle but statistically significant dynamic changes caused by emotional fluctuations, stress responses, etc. from a physiological level, significantly improving the objectivity, sensitivity and real-time nature of mental health monitoring, and providing solid technical support for personalized and intelligent mental health management.

[0042] (2) The present invention innovatively constructs a dynamic and hierarchical mental health early warning mechanism, which realizes continuous monitoring of the user's mental state and intelligent graded early warning based on the fusion analysis results of multi-dimensional physiological characteristics. Specifically, the system comprehensively evaluates the response intensity of the autonomic nervous system reflected by skin temperature fluctuations and the behavioral coordination represented by the deviation of body movement rhythm, uses a machine learning model to output a psychological stability score, and combines the time dimension to make a continuous judgment on the abnormal state; when the psychological stability score is lower than the set threshold, the system implements a graded early warning strategy based on the number of abnormal cycles: if it is a single cycle abnormality, the first-level early warning is activated and preliminary marking and mild intervention prompts are made; if it is in an unstable state for two or more consecutive cycles, the second-level early warning is triggered, and serious warning information is automatically pushed and the user is advised to accept professional mental health assessment or active intervention measures. This mechanism not only realizes the early identification and dynamic tracking of psychological risk events, but also effectively improves the practicality and personalization level of the early warning system. It can be widely used in wearable devices, mobile health platforms and telemedicine systems, and has significant social value and commercial prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The present invention will be further described below with reference to the accompanying drawings.

[0044] Figure 1 It is a flowchart of the intelligent mental health early warning method of the present invention;

[0045] Figure 2 It is a flowchart of the intelligent-based mental health early warning system in the present invention. DETAILED DESCRIPTION

[0046] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0047] See also Figure 1 As shown, the present invention is an intelligent-based mental health early warning method, comprising the following steps:

[0048] S1: During the mental health monitoring period, the user's skin temperature data and body movement frequency data are collected in real time;

[0049] S2: Obtain the user's skin temperature fluctuation data and calculate the differential temperature characteristic value based on the abnormal degree of skin differential temperature fluctuation to evaluate the intensity of the autonomic nervous system response under emotion induction;

[0050] S3: Analyze the rhythm stability of body movement frequency, calculate the characteristic value of body movement rhythm deviation, and evaluate the user's psychological coordination;

[0051] S4: Perform fusion analysis on the differential temperature characteristic value and the body movement rhythm deviation characteristic value, and judge whether the current psychological state is at an unstable level based on the analysis results;

[0052] S5: If it is determined to be a psychologically unstable state, the system will automatically activate the early warning mechanism.

[0053] In S1, during the mental health monitoring period, the user's skin temperature data and body movement frequency data are collected in real time, including:

[0054] In the mental health monitoring method of the present invention, a wearable physiological sensing device is first used to non-intrusively collect physiological data from the user during a set mental health monitoring period (e.g., every hour). Specifically, at the beginning of the monitoring period, the infrared sensing module and triaxial accelerometer integrated into the wearable device are activated to obtain the user's skin temperature profile and physical activity, respectively. Skin temperature data is continuously recorded at a set sampling frequency (e.g., once per second), forming a time series of temperature fluctuation data, thereby reflecting changes in the user's autonomic nervous system state.

[0055] Body movement frequency data is collected in multiple dimensions using a three-axis accelerometer, including the user's acceleration values in the X, Y, and Z directions. A built-in algorithm converts this raw data into body movement frequency and exercise intensity indicators per unit time. All collected data is transmitted in real time to a local processing unit for preliminary filtering and feature extraction, and then uploaded to a cloud server via a wireless communication module for subsequent mental health status assessment, trend analysis, and abnormality warnings. The entire collection process requires no active user intervention, ensuring data authenticity and continuity while protecting user privacy and data security.

[0056] In S2, the user's skin temperature fluctuation data is obtained. Based on the abnormal degree of skin differential temperature fluctuation, the differential temperature characteristic value is calculated to evaluate the intensity of the autonomic nervous system response under emotion induction, including:

[0057] During the mental health monitoring period, the user's skin temperature fluctuation data is obtained, and the differential temperature characteristic value is calculated based on the abnormal degree of skin differential temperature fluctuation. It is determined whether the differential temperature characteristic value is greater than or equal to the preset threshold. If so, the autonomic nervous system response under emotion induction is abnormal. If not, the autonomic nervous system response under emotion induction is normal.

[0058] The process of obtaining the differential temperature characteristic value is as follows:

[0059] During the mental health monitoring period, the user's skin temperature fluctuation data is collected to form the original temperature time series;

[0060] Calculate the absolute value of the temperature difference between adjacent original temperature times, and integrate the obtained absolute value of the temperature difference into a differential temperature data sequence;

[0061] Perform a Haar wavelet transform on the differential temperature data sequence to obtain the detail coefficient sequence and calculate the energy value of the detail coefficient. The calculation expression is: , where Represents the energy value of the detail coefficient, represents the number of detail coefficients, represents the total number of detail coefficients, Indicates the Detail coefficient;

[0062] The average energy value of all detail coefficients is calculated by using the mean calculation expression, and the ratio of the average energy value of all detail coefficients to the total energy value of all detail coefficients is calculated to obtain the differential temperature characteristic value.

[0063] It should be noted that: in the process of mental health monitoring, the present invention realizes a quantitative assessment of the response intensity of the autonomic nervous system under emotion induction by differential processing of the user's skin temperature fluctuation data and extracting the energy characteristics of the detail coefficients by combining the Haar wavelet transform. First, a time series of skin temperature is constructed, and a differential temperature series is formed by calculating the absolute value of the temperature difference between adjacent time points. Then, the Haar wavelet transform is used to extract its high-frequency change characteristics, and the ratio of the average energy to the total energy is calculated based on the energy distribution of the detail coefficient as the differential temperature characteristic value. The differential temperature characteristic value reflects the instantaneous fluctuation intensity in the process of body temperature regulation, thereby judging whether the activation state of the autonomic nervous system is abnormal. Compared with the traditional static threshold judgment method, the present invention has higher sensitivity and specificity, and is particularly suitable for identifying weak but continuous physiological changes caused by emotional fluctuations, thereby improving the accuracy and real-time performance of mental health status assessment.

[0064] In S3, the rhythm stability of body movement frequency is analyzed, the characteristic value of body movement rhythm deviation is calculated, and the user's psychological coordination is evaluated, including:

[0065] During the mental health monitoring period, the user's body movement frequency data is obtained, and the body movement rhythm deviation characteristic value is calculated based on the degree of deviation of the user's body movement frequency. It is determined whether the body movement rhythm deviation characteristic value is greater than or equal to the preset threshold. If so, the user's psychological coordination is abnormal; if not, the user's psychological coordination is normal.

[0066] The process of obtaining the body movement rhythm deviation characteristic value is as follows:

[0067] During the mental health monitoring period, the user's body movement frequency data is collected to form an original body movement time series;

[0068] Perform fast Fourier transform on the body motion frequency time series to obtain the frequency domain signal. Extract the main frequency components in the frequency domain signal. The calculation expression is: , , where Indicates the main frequency component of the current body motion pattern, represents the number of frequency components, Indicates the The complex amplitude of the frequency component, Indicates that the spectrum amplitude The maximum frequency , Indicates the The absolute value of the amplitude of the frequency component, Indicates the amplitude value at the corresponding frequency;

[0069] The absolute value of the difference between the main frequency component and the preset expected frequency is calculated and the ratio is calculated with the maximum frequency component to obtain the body movement rhythm deviation characteristic value.

[0070] It should be noted that: the present invention collects the user's body movement frequency data and performs frequency domain analysis on it, and proposes a method for calculating the characteristic value of body movement rhythm deviation based on fast Fourier transform, which is used to evaluate the behavioral coordination under the user's psychological state. First, the original body movement time series is converted into a frequency domain signal, and the main frequency components therein are identified, and compared with the preset expected frequency, and the degree of deviation is calculated to quantify the stability of the user's body movement rhythm. The characteristic value of body movement rhythm deviation can effectively reflect the degree of disorder of an individual's behavioral rhythm under emotional fluctuations or stress, and then determine whether his or her psychological coordination is abnormal. Compared with traditional methods based on time domain statistics, the present invention has higher sensitivity and robustness, can more accurately identify behavioral pattern deviations caused by changes in psychological state, and improves the comprehensive evaluation capability of the mental health monitoring system.

[0071] In S4, the differential temperature characteristic value and the body movement rhythm deviation characteristic value are fused and analyzed. Based on the analysis results, it is determined whether the current psychological state is at an unstable level, including:

[0072] During the mental health monitoring cycle, the user's psychological differential temperature characteristic values and body movement rhythm deviation characteristic values are obtained, and the differential temperature characteristic values and body movement rhythm deviation characteristic values are constructed into a comprehensive characteristic vector as the input of the mental health prediction model. The mental health prediction model is trained, and based on the trained mental health prediction model, the user's psychological stability score is output. Based on the user's psychological stability score, whether the current user's psychological state is stable is evaluated. The mental health prediction model is a gradient boosting tree model.

[0073] The training process of the mental health prediction model is as follows:

[0074] The comprehensive feature vector is obtained and input into the gradient boosting tree model to minimize the error between the predicted user psychological stability score and the actual user psychological stability score. This is used as the training target of the gradient boosting tree model. A cross-validation strategy is adopted, combined with grid search to optimize the hyperparameter configuration to improve the generalization ability and prediction accuracy of the model. The gradient descent method is used to minimize the loss function, and the decision tree set is iteratively updated to gradually enhance the model's ability to distinguish between psychologically stable and unstable states, thereby outputting an accurate user psychological stability score.

[0075] The determination of whether the current mental state is at an unstable level specifically includes:

[0076] During the mental health monitoring period, it is determined whether the psychological stability score of the current mental state is greater than or equal to a preset threshold. If so, the current user's mental state is at a stable level; if not, the current user's mental state is at an unstable level.

[0077] In S5, if the person is determined to be in a state of psychological instability, the system will automatically activate the early warning mechanism, including:

[0078] According to the unstable level of the current user's mental state, the number of cycles of the user's mental state instability is obtained. If the number of cycles of the user's mental state instability is less than two monitoring cycles, the user's mental state is determined to be a first-level abnormal state, and the current user is marked and a mental health warning is issued. If the number of cycles of the user's mental state instability is greater than or equal to two monitoring cycles, the user's mental state is determined to be a second-level abnormal state, and a mental health check is immediately performed on the current user, and a serious warning is issued.

[0079] See also Figure 2 As shown, the intelligent mental health early warning system includes:

[0080] A data acquisition module, which collects the user's skin temperature data and body movement frequency data in real time during the mental health monitoring period;

[0081] a user autonomic nervous system response evaluation module, which obtains the user's skin temperature fluctuation data and calculates a differential temperature characteristic value based on the abnormal degree of skin differential temperature fluctuation, for use in evaluating the intensity of the autonomic nervous system response under emotion induction;

[0082] A user psychological coordination evaluation module, which evaluates the user's psychological coordination by analyzing the rhythm stability of body movement frequency and calculating the body movement rhythm deviation characteristic value;

[0083] A user psychological state stability assessment module, which performs a fusion analysis on the differential temperature characteristic value and the body movement rhythm deviation characteristic value, and determines whether the current psychological state is at an unstable level based on the analysis result;

[0084] The early warning module automatically activates the early warning mechanism if the early warning module determines that the patient is in a state of psychological instability.

[0085] The present invention operates by dynamically monitoring a user's psychological state and providing precise early warnings through the real-time acquisition and fusion analysis of multimodal physiological signals. During the psychological health monitoring cycle, the present invention utilizes the infrared sensor and triaxial accelerometer in the wearable device to continuously collect data on the user's skin temperature fluctuations and body movement frequency, reflecting the activation intensity of the autonomic nervous system and the coordination of behavioral rhythms, respectively. By performing differential processing on the skin temperature series and extracting high-frequency energy features using the Haar wavelet transform, differential temperature eigenvalues are constructed to quantify physiological fluctuations induced by emotions. Simultaneously, the body movement frequency series is modeled in the frequency domain using a fast Fourier transform, and the body movement rhythm deviation eigenvalue is calculated to assess the degree of alignment between the individual's behavioral rhythm and the ideal pattern. Furthermore, these two features are fused and input into a trained gradient boosting tree model to intelligently predict the user's psychological stability. The prediction results determine whether to trigger an early warning mechanism. The system then sets a graded early warning strategy based on the duration of the abnormal state, improving the timeliness and targeted nature of intervention. This invention integrates physiological signal processing, machine learning and psychophysiological principles, breaking through the limitations of traditional subjective assessment methods. It has the advantages of being non-invasive, high-precision and highly real-time, and is suitable for personalized mental health management and remote early warning services under smart wearable platforms.

[0086] The above is a detailed description of an embodiment of the present invention. However, the content described is only a preferred embodiment of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. An intelligent mental health early warning method, characterized in that: The following steps are involved: S1: During the mental health monitoring period, the user's skin temperature data and body movement frequency data are collected in real time; S2: Obtain the user's skin temperature fluctuation data and calculate the differential temperature characteristic value based on the abnormal degree of skin differential temperature fluctuation to evaluate the intensity of the autonomic nervous system response under emotion induction; S3: Analyze the rhythm stability of body movement frequency, calculate the characteristic value of body movement rhythm deviation, and evaluate the user's psychological coordination; S4: Perform fusion analysis on the differential temperature characteristic value and the body movement rhythm deviation characteristic value, and judge whether the current psychological state is at an unstable level based on the analysis results; S5: If it is determined to be a psychologically unstable state, the system will automatically activate the early warning mechanism.

2. The intelligent mental health early warning method according to claim 1, characterized in that: The assessment of the intensity of the autonomic nervous system response under emotion induction specifically includes: During the mental health monitoring period, the user's skin temperature fluctuation data is obtained, and the differential temperature characteristic value is calculated based on the abnormal degree of skin differential temperature fluctuation. It is determined whether the differential temperature characteristic value is greater than or equal to the preset threshold. If so, the autonomic nervous system response under emotion induction is abnormal. If not, the autonomic nervous system response under emotion induction is normal.

3. The intelligent mental health early warning method according to claim 2, characterized in that: The process of obtaining the differential temperature characteristic value is as follows: During the mental health monitoring period, the user's skin temperature fluctuation data is collected to form the original temperature time series; Calculate the absolute value of the temperature difference between adjacent original temperature times, and integrate the obtained absolute value of the temperature difference into a differential temperature data sequence; Perform a Haar wavelet transform on the differential temperature data sequence to obtain a detail coefficient sequence and calculate the energy value of the detail coefficient; The average energy value of all detail coefficients is calculated by using the mean calculation expression, and the ratio of the average energy value of all detail coefficients to the total energy value of all detail coefficients is calculated to obtain the differential temperature characteristic value.

4. The intelligent mental health early warning method according to claim 1, characterized in that: The evaluation of user psychological coordination specifically includes: During the mental health monitoring period, the user's body movement frequency data is obtained, and the body movement rhythm deviation characteristic value is calculated based on the degree of deviation of the user's body movement frequency. It is determined whether the body movement rhythm deviation characteristic value is greater than or equal to the preset threshold. If so, the user's psychological coordination is abnormal; if not, the user's psychological coordination is normal.

5. The intelligent-based mental health early warning method according to claim 4, characterized in that: The process of obtaining the body movement rhythm deviation characteristic value is as follows: During the mental health monitoring period, the user's body movement frequency data is collected to form an original body movement time series; Perform fast Fourier transform on the body motion frequency time series to obtain the frequency domain signal, and extract the main frequency components in the frequency domain signal; The absolute value of the difference between the main frequency component and the preset expected frequency is calculated and the ratio is calculated with the maximum frequency component to obtain the body movement rhythm deviation characteristic value.

6. The intelligent mental health early warning method according to claim 1, characterized in that: The fusion analysis of the differential temperature characteristic value and the body movement rhythm deviation characteristic value specifically includes: During the mental health monitoring cycle, the user's psychological differential temperature characteristic values and body movement rhythm deviation characteristic values are obtained, and the differential temperature characteristic values and body movement rhythm deviation characteristic values are constructed into a comprehensive characteristic vector as the input of the mental health prediction model. The mental health prediction model is trained, and based on the trained mental health prediction model, the user's psychological stability score is output. Based on the user's psychological stability score, whether the current user's psychological state is stable is evaluated. The mental health prediction model is a gradient boosting tree model.

7. The intelligent-based mental health early warning method according to claim 6, characterized in that: The training process of the mental health prediction model is as follows: The comprehensive feature vector is obtained and input into the gradient boosting tree model to minimize the error between the predicted user psychological stability score and the actual user psychological stability score. This is used as the training target of the gradient boosting tree model. A cross-validation strategy is adopted, combined with grid search to optimize the hyperparameter configuration to improve the generalization ability and prediction accuracy of the model. The gradient descent method is used to minimize the loss function, and the decision tree set is iteratively updated to gradually enhance the model's ability to distinguish between psychologically stable and unstable states, thereby outputting an accurate user psychological stability score.

8. The intelligent-based mental health early warning method according to claim 1, characterized in that: The determination of whether the current mental state is at an unstable level specifically includes: During the mental health monitoring period, it is determined whether the psychological stability score of the current mental state is greater than or equal to a preset threshold. If so, the current user's mental state is at a stable level; if not, the current user's mental state is at an unstable level.

9. The intelligent-based mental health early warning method according to claim 1, characterized in that: If the patient is determined to be in a state of psychological instability, the system will automatically activate the early warning mechanism, including: According to the unstable level of the current user's mental state, the number of cycles of the user's mental state instability is obtained. If the number of cycles of the user's mental state instability is less than two monitoring cycles, the user's mental state is determined to be a first-level abnormal state, and the current user is marked and a mental health warning is issued. If the number of cycles of the user's mental state instability is greater than or equal to two monitoring cycles, the user's mental state is determined to be a second-level abnormal state, and a mental health check is immediately performed on the current user, and a serious warning is issued.

10. An intelligent-based mental health early warning system, used for the intelligent-based mental health early warning system according to any one of claims 1 to 9, characterized in that: include: A data acquisition module, which collects the user's skin temperature data and body movement frequency data in real time during the mental health monitoring period; a user autonomic nervous system response evaluation module, which obtains the user's skin temperature fluctuation data and calculates a differential temperature characteristic value based on the abnormal degree of skin differential temperature fluctuation, for use in evaluating the intensity of the autonomic nervous system response under emotion induction; A user psychological coordination evaluation module, which evaluates the user's psychological coordination by analyzing the rhythm stability of body movement frequency and calculating the body movement rhythm deviation characteristic value; A user psychological state stability assessment module, which performs a fusion analysis on the differential temperature characteristic value and the body movement rhythm deviation characteristic value, and determines whether the current psychological state is at an unstable level based on the analysis result; The early warning module automatically activates the early warning mechanism if the early warning module determines that the patient is in a state of psychological instability.

Images