AI mental health monitoring method and device based on traditional chinese medicine data and electronic equipment

Abstract

The application provides an AI mental health monitoring method and device based on traditional Chinese medicine data and electronic equipment, and relates to the field of monitoring. The method comprises the following steps: collecting pulse data of a user through a pulse sensor, and collecting electroencephalogram data of the user through an electroencephalogram collector; in response to at least two kinds of real-time change data in 3D dynamic model, pulse data and electroencephalogram data meeting specified psychological emotional fluctuation phenomenon data, judging whether the interval of the change occurrence time corresponding to the at least two kinds of real-time change data is less than a specified time interval, if the interval is less than the specified time interval, determining the surrounding environment voice content and the surrounding environment image content corresponding to the change occurrence time from the surrounding environment video, analyzing the psychological stimulation point object of the user based on the surrounding environment voice content and the surrounding environment image content through an AI system, and generating a mental health monitoring strategy for the user according to a plurality of psychological stimulation point objects.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of monitoring, in particular to an AI mental health monitoring method and device based on traditional Chinese medicine data and electronic equipment. BACKGROUND

[0002] At present, traditional psychological assessment relies on questionnaire survey and interview, and has the problems of low efficiency, narrow coverage and difficulty in continuous monitoring. Although the existing intelligent mental health management system combines intelligent hardware, cloud computing and other technologies, it lacks the systematic integration of the diagnosis thought of traditional Chinese medicine of looking, listening, asking and palpation, so that the accuracy of the psychological monitoring result of the user is low, resulting in low accuracy of the mental health monitoring of the user. SUMMARY

[0003] The purpose of the present application is to provide an AI mental health monitoring method and device based on traditional Chinese medicine data to solve the technical problem of low accuracy of mental health monitoring of the user.

[0004] In a first aspect, the present application provides an AI mental health monitoring method based on traditional Chinese medicine data, applied to a monitoring system, wherein the monitoring system is connected with a pulse sensor, a camera and an electroencephalogram collector, the camera is arranged on a drone, and an ultra-wideband chip is arranged on the camera; the method comprises:

[0005] The pulse data of the user is collected by the pulse sensor, and the electroencephalogram data of the user is collected by the electroencephalogram collector;

[0006] The drone is controlled to fly to a plurality of different positions around the user according to the real-time positioning of the ultra-wideband chip, a plurality of different relative position relationships of the plurality of positions relative to the user are determined by the ultra-wideband chip, dynamic images of the user and surrounding environment videos corresponding to the time period of the dynamic images are obtained at each position by the camera based on the plurality of relative position relationships, and a 3D dynamic model of the user is generated based on the plurality of relative position relationships and the corresponding collected dynamic images; the dynamic images include facial variation images, body posture change images and action execution images of the user;

[0007] In response to at least two real-time variation data of the 3D dynamic model, the pulse data and the electroencephalogram data meeting the specified psychological emotional fluctuation phenomenon data, it is judged whether the interval of the variation occurrence time corresponding to the at least two real-time variation data is less than a specified time interval, if less than the specified time interval, the surrounding environment sound content and the surrounding environment image content corresponding to the variation occurrence time are determined from the surrounding environment video, the psychological stimulus point object of the user is analyzed by the AI system based on the surrounding environment sound content and the surrounding environment image content, and the mental health monitoring strategy for the user is generated according to a plurality of psychological stimulus point objects.

[0008] In a possible implementation, the pulse sensor is arranged on a whole pulse detection wristband, an ultrasonic radar measurer and an air bag are arranged on the inner side of the whole pulse detection wristband, the ultrasonic radar measurer is arranged inside the air bag close to the wristband and has a measurement direction towards the inner side of the wristband, and the film thickness of the air bag is less than a specified thickness; the psychological health monitoring strategy for the user is generated according to the psychological stimulation point objects, and the psychological health monitoring strategy comprises the following steps:

[0009] The pulse frequency and the pulse strength of the user are detected by the pulse sensor in a piezoelectric manner;

[0010] When the deformation degree caused by the inflation amount of the air bag meets a specified deformation degree and stops deforming again, the dynamic pulse overall fluctuation amplitude and the dynamic pulse overall fluctuation form of the user are determined according to the distance change data measured by the ultrasonic radar measurer, wherein the distance change data comprises the distance change conditions between a plurality of skin position points in the overall skin area range corresponding to the wrist pulse range of the user measured by the ultrasonic radar measurer;

[0011] The dynamic pulse overall simulation 3D model of the user is generated based on the dynamic pulse overall fluctuation form, the dynamic pulse overall fluctuation amplitude, the pulse frequency and the pulse strength;

[0012] The psychological health monitoring strategy for the user is generated according to the psychological stimulation point objects and the target dynamic pulse overall simulation 3D model corresponding to the occurrence time of each psychological stimulation point object.

[0013] In a possible implementation, after the dynamic pulse overall simulation 3D model of the user is generated based on the dynamic pulse overall fluctuation form, the pulse frequency and the pulse strength, the following step is further included:

[0014] The dynamic pulse overall simulation 3D model, the plurality of skin position points and the distance change data corresponding to each skin position point are transmitted to a 3D pulse dynamic simulator, wherein a plurality of nail-shaped protrusion points are arranged on the surface of the 3D pulse dynamic simulator;

[0015] The target nail-shaped protrusion point corresponding to the position of the plurality of skin position points is determined from the plurality of nail-shaped protrusion points;

[0016] Based on the overall dynamic pulse simulation 3D model and the distance change data corresponding to each of the target nail-shaped protrusion points, the simulation HD vibrator in the pulse dynamic simulator is used to control the protrusion time, protrusion position, and protrusion speed of the target nail-shaped protrusion points. The overall dynamic pulse fluctuation of the user is simulated by the protrusion time, protrusion position, and protrusion speed of the target nail-shaped protrusion points, so as to perform TCM pulse diagnosis based on the overall dynamic pulse fluctuation.

[0017] In one possible implementation, the step of controlling the protrusion time, protrusion position, and protrusion velocity of the target nail-shaped protrusions using a simulated HD vibrator in the pulse dynamic simulator, based on the overall 3D simulation model of the dynamic pulse and the distance change data corresponding to each of the target nail-shaped protrusions, includes:

[0018] According to the specified magnification ratio, the overall fluctuation amplitude and pulse strength of the dynamic pulse in the overall simulation 3D model of the dynamic pulse are magnified while keeping the overall fluctuation shape and pulse frequency of the dynamic pulse unchanged, so as to generate the user's dynamic pulse magnification simulation highlighting model.

[0019] Based on the dynamic pulse amplification simulation protrusion model and the amplified distance change data corresponding to each of the target nail-shaped protrusion points, the protrusion time, protrusion position, and protrusion speed of the target nail-shaped protrusion points are controlled by the simulated HD vibrator in the pulse dynamic simulator.

[0020] In one possible implementation, a pressure sensor is provided on the 3D pulse dynamic simulator; the method further includes:

[0021] In response to the pulse diagnosis operation applied to the 3D pulse dynamic simulator, the pressure sensor detects the operation pressure data corresponding to the pulse diagnosis operation.

[0022] The operating pressure data is transmitted to the airbag controller corresponding to the airbag.

[0023] The airbag controller controls the airbag to simulate the operating pressure of the pulse diagnosis operation by adjusting the pressure generated by different airbag inflation volumes according to the operating pressure data.

[0024] In one possible implementation, the step of controlling the airbag via the airbag controller to simulate the operating pressure of the pulse diagnosis operation by adjusting the pressure generated by different airbag inflation volumes according to the operating pressure data includes:

[0025] Based on the operating pressure of the pulse diagnosis operation, the elastic coefficient of the airbag, and the contact area between the airbag and the user's skin, the target inflation volume of the airbag to be inflated is determined using the following formula, so that the airbag pressure generated by the target inflation volume can simulate the operating pressure of the pulse diagnosis operation:

[0026]

[0027] in, This indicates the inflation volume within the target airbag; This indicates the operating pressure of the pulse diagnosis operation; This indicates the contact area between the airbag and the user's skin; This represents the ideal gas constant within the airbag; This indicates the ambient temperature of the environment in which the airbag is located; This indicates the atmospheric pressure of the environment in which the airbag is located; This indicates the initial inflation volume within the airbag; The elastic coefficient of the airbag is an equivalent stiffness parameter used to characterize the ability of the thin film material of the airbag to resist deformation under pressure.

[0028] Based on the target air volume, the airbag controller controls the inflation of the airbag.

[0029] In one possible implementation, the monitoring system is also connected to a tongue scanner, a respiratory flow sensor, and a taste sensor; the step of generating a mental health monitoring strategy for the user based on a 3D simulation model of the overall dynamic pulse corresponding to the occurrence time of each of the aforementioned psychological stimulus points includes:

[0030] The tongue image scanner collects the user's tongue image data, the respiratory flow sensor converts the user's inhaled and exhaled gas flow into electrical signals and detects the user's ventilation per unit time, breath flow rate parameters, tidal volume, breath temperature and oxygen and carbon dioxide concentrations based on the electrical signals, and the taste sensor collects the user's body odor data.

[0031] Based on the dynamic pulse overall simulation 3D model, the 3D dynamic model, the electroencephalogram data, the tongue data, the ventilation per unit time, the breath flow rate parameter, the tidal volume, the breath temperature, the body odor data, and the oxygen and carbon dioxide concentration, the AI ​​system analyzes the user's visceral pathological data.

[0032] Based on the data on visceral lesions, several psychological stimulation points, and a 3D simulation model of the target dynamic pulse corresponding to the occurrence time of each psychological stimulation point, a mental health monitoring plan is generated for the user.

[0033] In one possible implementation, the EEG acquisition device is a multi-channel EEG acquisition device; the specified psychological and emotional fluctuation phenomenon data includes at least one of the following:

[0034] Data on changes in body posture, movements, facial expressions, pulse, and various types of electroencephalograms (EEGs) corresponding to fluctuations in psychological and emotional states.

[0035] Secondly, this application provides an AI-based mental health monitoring device based on traditional Chinese medicine data, applied to a monitoring system. The monitoring system is connected to a pulse sensor, a camera, and an electroencephalogram (EEG) collector. The camera is mounted on a drone and is equipped with an ultra-wideband chip. The device includes:

[0036] The acquisition module is used to acquire the user's pulse data through the pulse sensor and to acquire the user's brainwave data through the brainwave acquisition device.

[0037] The generation module is used to control the drone to fly to multiple different locations around the user based on the real-time positioning of the ultra-wideband chip, determine multiple different relative positional relationships between the multiple locations and the user using the ultra-wideband chip, acquire dynamic images of the user and surrounding environment videos of the corresponding time period through the camera at each location, and generate a 3D dynamic model of the user based on the multiple relative positional relationships and the corresponding acquired dynamic images; the dynamic images include images of facial changes, body posture changes, and action execution.

[0038] The analysis module is used to respond to at least two real-time changes in the 3D dynamic model, the pulse data, and the EEG data that conform to specified psychological and emotional fluctuation phenomena. It determines whether the interval between the occurrence times of the changes corresponding to the at least two real-time changes is less than a specified time interval. If it is less than the specified time interval, it determines the surrounding environment audio content and surrounding environment image content corresponding to the occurrence time of the change from the surrounding environment video. Based on the surrounding environment audio content and surrounding environment image content, it analyzes the user's psychological stimulus points through the AI ​​system and generates a psychological health monitoring strategy for the user based on several of the psychological stimulus points.

[0039] Thirdly, this application also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the computer program to implement the method described in the first aspect above.

[0040] Fourthly, this application also provides a computer-readable storage medium storing computer-executable instructions that, when invoked and executed by a processor, cause the processor to perform the method described in the first aspect above.

[0041] This application provides an AI-based mental health monitoring method, device, and electronic device based on Traditional Chinese Medicine (TCM) data, applied to a monitoring system. The monitoring system is connected to a pulse sensor, a camera, and an EEG acquisition device. The camera is mounted on a drone and equipped with an ultra-wideband (UWB) chip. The method can collect the user's pulse data via the pulse sensor and the user's EEG data via the EEG acquisition device. Based on the real-time positioning control of the UWB chip, the drone flies to multiple different locations around the user. The UWB chip determines multiple relative positional relationships between these locations and the user. Based on each location, the camera acquires a dynamic image of the user and a corresponding time period of the surrounding environment video. The method then uses these multiple relative positional relationships to... The system generates a 3D dynamic model of the user based on the collected dynamic images. The dynamic images include images of facial changes, body posture changes, and action execution. In response to at least two real-time changes in the 3D dynamic model, pulse data, and EEG data conforming to specified psychological and emotional fluctuation phenomena, the system determines whether the interval between the occurrence times of the changes corresponding to the at least two real-time changes is less than a specified time interval. If it is less than the specified time interval, the system determines the surrounding environment audio content and surrounding environment image content corresponding to the occurrence time of the change from the surrounding environment video. Based on the surrounding environment audio content and surrounding environment image content, the system analyzes the user's psychological stimulus points through the AI ​​system and generates a psychological health monitoring strategy for the user based on several of the psychological stimulus points.This solution captures emotional fluctuations from both physiological and behavioral dimensions by simultaneously acquiring pulse, EEG, and 3D dynamic behavior models. This multi-source data fusion enhances the reliability of emotion recognition and avoids misjudgments based on a single modality. Furthermore, a UWB chip-equipped drone achieves centimeter-level positioning, dynamically acquiring user images from multiple perspectives to reconstruct a high-fidelity 3D dynamic model. This comprehensively restores micro-expressions and subtle movements, achieving high-precision spatial modeling and improving the granularity of behavior analysis. It overcomes the problems of occlusion and information loss caused by fixed camera perspectives. Moreover, an emotional event is triggered only when at least two modalities of data synchronously exhibit changes consistent with emotional characteristics within a specified time interval. This effectively eliminates accidental noise or isolated anomalies, achieving temporal consistency verification and reducing false alarm rates. If the emotion is confirmed... The event is immediately linked to the corresponding environmental audio and video at that moment. AI analysis identifies specific psychological stimuli, such as specific people, words, or scenes, advancing monitoring from state recognition to causal tracing. This allows for precise location of the stimulus source through environmental context backtracking. Based on historical stimuli and user response patterns, targeted monitoring and intervention strategies are dynamically generated, improving the system's long-term adaptability and intervention effectiveness. This personalized strategy generation forms a closed-loop intervention. Therefore, through the aforementioned multimodal perception, spatiotemporal alignment, environmental correlation, and causal inference technologies, more accurate, in-depth, and interpretable monitoring of users' psychological states is achieved, significantly improving the accuracy of mental health monitoring and solving the technical problem of low accuracy in monitoring users' mental health.

[0042] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the specific embodiments of this application or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0044] Figure 1 A flowchart illustrating the AI-based mental health monitoring method based on traditional Chinese medicine data provided in this application embodiment;

[0045] Figure 2 Another flowchart illustrating the AI-based mental health monitoring method based on traditional Chinese medicine data provided in this application embodiment;

[0046] Figure 3 A schematic diagram of the structure of an AI-based mental health monitoring device based on traditional Chinese medicine data provided in this application embodiment;

[0047] Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of this application is shown. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0049] The terms "comprising" and "having," and any variations thereof, used in the embodiments of this application, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the steps or units listed, but may optionally include other steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0050] This application provides an AI-based mental health monitoring method, device, and electronic device based on traditional Chinese medicine data. This method can solve the technical problem of low accuracy in monitoring users' mental health.

[0051] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0052] Figure 1 This is a flowchart illustrating an AI-based mental health monitoring method based on Traditional Chinese Medicine (TCM) data, provided as an embodiment of this application. The method is applied to a monitoring system, which is connected to a pulse sensor, a camera, and an electroencephalogram (EEG) scanner. The camera is mounted on a drone and is equipped with an ultra-wideband chip. Figure 1 As shown, the method includes:

[0053] Step S110: Collect the user's pulse data through the pulse sensor and collect the user's brainwave data through the EEG collector.

[0054] For example, the system automatically detects and initializes paired pulse sensors and EEG acquisition devices (e.g., via Bluetooth, Wi-Fi, or a dedicated interface). Then, it verifies the user's identity (e.g., through biometrics, account login, or device wearing confirmation), establishing a binding relationship between the current session and the specific user. Afterward, the monitoring system's main control module sends a synchronization sampling command to ensure that the pulse sensor and EEG acquisition device begin data acquisition at the same time reference, supporting subsequent multimodal timing alignment. Hardware synchronization signals (e.g., GPIO triggering) or software timestamp alignment mechanisms (e.g., NTP / PTP protocols) can be used. The pulse sensor (e.g., a photoplethysmography (PPG) sensor) detects blood flow changes in areas such as the user's fingertips, wrist, or earlobe in real time, outputting raw analog or digital signals. After analog-to-digital conversion (ADC), the signal generates time-series pulse waveform data, including potential features such as heart rate (HR) and heart rate variability (HRV). The EEG acquisition device (e.g., a dry electrode-based EEG headset) then acquires scalp potential differences through multiple channels, obtaining weak electrical signals at the μV level. After amplification, filtering (to remove power frequency interference, electromyographic noise, etc.), and analog-to-digital conversion, the signal is converted into multi-channel time-voltage sequence data. Both types of data are stamped with a high-precision, unified timestamp (e.g., at the microsecond level) at the acquisition or receiving end to ensure temporal consistency in subsequent fusion analysis. The raw data is temporarily stored in a local buffer or edge computing node, awaiting uploading or real-time processing.

[0055] Step S120: Based on the real-time positioning control of the ultra-wideband chip, the drone flies to multiple different locations around the user. The ultra-wideband chip is used to determine multiple different relative positional relationships between the multiple locations and the user. Based on each location, the camera acquires dynamic images of the user and surrounding environment videos for the corresponding time period of the dynamic images. Based on the multiple relative positional relationships and the corresponding acquired dynamic images, a 3D dynamic model of the user is generated.

[0056] The dynamic images include images of the user's facial changes, body posture changes, and actions performed.

[0057] As an optional implementation, a UWB tag deployed on the user (or integrated into a wearable device) establishes communication with several UWB anchor points in the environment (or the drone itself acts as a mobile anchor point). Using TDOA (Time Difference of Arrival) or TOF (Time of Flight) algorithms, the user's precise coordinates in three-dimensional space are calculated in real time (accuracy down to the centimeter level). These coordinates serve as the reference origin for subsequent drone path planning. Then, the monitoring system's main control module automatically generates a set of target flight positions around the user based on preset strategies (such as spherical sampling, view frustum coverage, and occlusion avoidance). Each target position meets the following requirements: maintaining a safe distance from the user (e.g., 1–3 meters); covering different azimuth and pitch angles (ensuring visibility of the face, torso, and limbs); and avoiding environmental obstacles (combined with SLAM or pre-built mapping information).

[0058] The UAV then receives flight commands and calculates its current position vector relative to the user in real time based on the ranging data between its own UWB module and the user's UWB tag. Utilizing an onboard flight control system (such as PX4 / ArduPilot) combined with UWB+IMU fusion positioning, it accurately flies to the target location and hovers stably. At each target location, the system records the precise relative positional relationship (including distance, azimuth, and pitch angle) and the corresponding timestamp. Simultaneously, at each hovering target location, the UAV performs the following operations: activates its camera and continuously acquires dynamic image sequences containing the user at a high frame rate (e.g., 30–60 fps), covering: facial micro-expression changes (e.g., frowning, twitching of the mouth); changes in body posture (e.g., hunching, trembling, restlessness); and the execution of actions (e.g., waving, pacing, holding head). It also simultaneously records video of the surrounding environment (including audio), recording the scene context (background people, objects, sounds, etc.) within the same time period. All video streams are stamped with high-precision timestamps consistent with the physiological data to ensure cross-modal alignment. The system aggregates dynamic image sequences from n different perspectives and their corresponding relative positional relationships. By utilizing the principle of multi-view geometry, 2D image features (such as key points, contours, and textures) from various perspectives are mapped to a unified world coordinate system (centered on the user).

[0059] Then, based on the fused multi-view data, a time-varying 3D dynamic model is generated using any one or a combination of the following techniques: dynamic reconstruction based on neural radiation fields (NeRF): constructing an implicit 3D representation that can render any viewpoint; parametric human model fitting (such as SMPL / SMPL-X): inferring 3D joint and deformation parameters from 2D keypoints; stereo vision + deep learning fusion: improving surface detail accuracy using disparity maps and semantic segmentation. The output is a time-series 3D mesh or skeleton model that accurately reflects the user's facial expression changes, body posture evolution, and movement trajectory during the monitoring period. The generated 3D dynamic model, along with the original multi-view video and environmental context data, is then fed into an AI analysis engine for: cross-modal emotional event detection with pulse and EEG data; identification of abnormal behavioral patterns (such as self-harm tendencies and panic attacks); and support for the tracing of psychological stimuli and the generation of intervention strategies.

[0060] Step S130: In response to at least two real-time change data in the 3D dynamic model, pulse data, and EEG data conforming to specified psychological and emotional fluctuation phenomena, determine whether the interval between the occurrence times of the changes corresponding to the at least two real-time change data is less than a specified time interval. If it is less than the specified time interval, determine the surrounding environment voice content and surrounding environment image content corresponding to the occurrence time of the change from the surrounding environment video. Based on the surrounding environment voice content and surrounding environment image content, analyze the user's psychological stimulus points through the AI ​​system, and generate a psychological health monitoring strategy for the user based on several psychological stimulus points.

[0061] As an optional implementation method, the EEG acquisition device is a multi-channel EEG acquisition device; the specified psychological and emotional fluctuation phenomenon data includes at least one of the following: changes in body posture, movement, facial expression, pulse, and various types of EEG waves corresponding to the occurrence of psychological and emotional fluctuations.

[0062] Multi-channel EEG can simultaneously capture the electrophysiological activity of different brain regions (such as the prefrontal cortex, temporal lobe, and parietal lobe), thereby extracting frequency domain features (such as alpha, beta, theta, and gamma wave power), spatial distribution patterns (such as alpha asymmetry between the left and right prefrontal lobes), and functional connectivity indicators closely related to emotions. These features are highly sensitive in distinguishing different emotions such as anxiety, depression, anger, and pleasure. Moreover, by jointly modeling multi-channel EEG changes with overt behavioral / physiological signals such as body posture, movement, facial expressions, and pulse, a cross-modal emotional feature vector can be formed. For example, an anxious state may be characterized by: enhanced theta waves in the prefrontal cortex, increased heart rate, slight frowning, and body stiffness; a pleasure state may be characterized by: alpha inhibition in the left prefrontal cortex, stable heart rate, upturned corners of the mouth, and open gestures. This multi-dimensional consistency matching significantly reduces the risk of misjudgment based on a single modality (such as "fake smile" without EEG support, or "sitting still but feeling panicked" with no facial expression but abnormal EEG / pulse).

[0063] In one possible implementation, a pulse sensor is mounted on a pulse wave detection wristband. An ultrasonic radar measuring device and an airbag are located on the inner side of the wristband. The ultrasonic radar measuring device is positioned inside the airbag, close to the wristband, and its measurement direction is towards the inner side of the wristband. The thickness of the airbag's film is less than a specified thickness. The aforementioned generation of a user-specific mental health monitoring strategy based on several psychological stimulus points may specifically include the following steps:

[0064] The pulse sensor uses piezoelectricity to detect the user's pulse frequency and pulse strength. When the deformation caused by the inflation volume in the airbag meets a specified deformation level and stops, the overall amplitude and morphology of the user's dynamic pulse are determined based on multiple distance change data measured by an ultrasonic radar. These multiple distance change data include the distance changes between multiple skin locations within the overall skin area corresponding to the user's wrist pulse range, as measured by the ultrasonic radar. A 3D simulation model of the user's dynamic pulse is generated based on the overall dynamic pulse morphology, amplitude, frequency, and strength. A mental health monitoring strategy for the user is generated based on several psychological stimulus points and the target dynamic pulse simulation model corresponding to the occurrence time of each stimulus point.

[0065] Traditional wearable devices can only measure heart rate (frequency) or a rough pulse strength, lacking the ability to capture the morphological characteristics of the pulse (such as its depth, slowness, rapidity, smoothness, roughness, and tightness). This solution, however, uses a piezoelectric sensor to acquire pulse frequency and strength (corresponding to "rapidity" and "strength"). Ultrasonic radar, combined with an ultra-thin airbag, measures the micron-level fluctuations at multiple points on the wrist skin surface with high precision in a non-contact manner, reconstructing the dynamic displacement field of the entire Cun-Kou area (radial artery distribution area). This achieves, for the first time, the digital reconstruction of the overall pulse amplitude and morphology (i.e., "pulse shape") in a wearable device, approximating the holistic view of the three divisions and nine pulse points in Traditional Chinese Medicine. For the first time, wearable devices enable high-precision digital acquisition and 3D modeling of the overall dynamic characteristics of the pulse in Traditional Chinese Medicine, linking it to psychological stimuli to generate a logically personalized mental health monitoring strategy that combines physiological depth, emotional specificity, and Traditional Chinese Medicine diagnostic principles.

[0066] As one possible implementation, the monitoring system is also connected to a tongue scanner, a respiratory flow sensor, and a taste sensor; the above-mentioned generation of a user's mental health monitoring strategy based on a 3D simulation model of several psychological stimulus points and the target dynamic pulse corresponding to the occurrence time of each psychological stimulus point can specifically include the following steps:

[0067] The system collects the user's tongue image data through a tongue image scanner, converts the user's inhaled and exhaled gas flow rate into electrical signals through a respiratory flow sensor, and detects the user's ventilation per unit time, breath flow rate parameters, tidal volume, breath temperature, and oxygen and carbon dioxide concentrations based on the electrical signals. The system also collects the user's body odor data through a taste sensor.

[0068] Based on the dynamic pulse overall simulation 3D model, 3D dynamic model, and EEG data, tongue data, ventilation per unit time, breath flow rate parameters, tidal volume, breath temperature, body odor data, and oxygen and carbon dioxide concentration, the AI ​​system analyzes the user's visceral pathological data.

[0069] Based on data on visceral pathological changes, several psychological stressors, and a 3D simulation model of the target dynamic pulse corresponding to the occurrence time of each psychological stressor, a mental health monitoring plan for the user is generated.

[0070] By integrating tongue scanners, respiratory flow sensors, and taste sensors, and fusing them with existing dynamic pulse 3D models, 3D behavioral models, and EEG data, a holistic TCM intelligent analysis system for mind-body integration has been constructed. This system enables cross-system correlation deduction from psychological and emotional representations to the functional state of internal organs, thereby generating a highly individualized and logically supported mental health monitoring program with TCM diagnostic support. It breaks through the limitations of traditional purely psychological mental health monitoring by introducing a physiological-organ dimension. The AI ​​mental health monitoring system fully integrates objective sensor data from the four diagnostic methods of TCM (tongue observation, taste olfaction, pulse diagnosis, and emotional inquiry). Through multimodal AI modeling, it achieves a dynamic mapping of psychological stimulation, emotional response, and organ function, generating a personalized, executable, and preventative mental health monitoring program that combines modern physiological precision with the wisdom of TCM's holistic perspective.

[0071] In some embodiments, after generating a 3D simulation model of the user's overall dynamic pulse based on the overall fluctuation pattern of the dynamic pulse, pulse frequency, and pulse strength, as follows: Figure 2 As shown, the method may further include the following steps:

[0072] Step S210: The dynamic pulse pattern overall simulation 3D model, multiple skin location points and the distance change data corresponding to each skin location point are transmitted to the 3D pulse pattern dynamic simulator; wherein, multiple nail-shaped protrusions are set on the surface of the 3D pulse pattern dynamic simulator.

[0073] Step S220: Determine the target nail-like protrusion point corresponding to the multiple skin location points from among the multiple nail-like protrusion points;

[0074] Step S230: Based on the dynamic pulse overall simulation 3D model and the distance change data corresponding to each target nail-shaped protrusion point, the simulation HD vibrator in the pulse dynamic simulator is used to control the protrusion time, protrusion position and protrusion speed of the target nail-shaped protrusion points. The overall fluctuation of the user's dynamic pulse is simulated by the protrusion time, protrusion position and protrusion speed of the target nail-shaped protrusion points, so as to perform TCM pulse diagnosis based on the overall fluctuation of the dynamic pulse.

[0075] In this embodiment, after generating a dynamic pulse pattern 3D model of the user, the pulse pattern is further physically reproduced and tactilely simulated using a 3D pulse pattern dynamic simulator (equipped with a programmable nail-shaped protrusion point and an HD vibrator). This achieves high-fidelity, tactile digital reproduction of the pulse pattern in traditional Chinese medicine, enabling AI systems or TCM doctors to perform remote, objective, and repeatable pulse diagnosis through machine pulse taking, breaking through the limitations of traditional pulse diagnosis that relies on subjective touch and face-to-face contact.

[0076] By mapping the model onto a physical surface with nail-like protrusions (simulating the skin at the cun-kou point), and using a high-precision HD vibrator to control the protrusion time, position, and speed of each protrusion (corresponding to the arrival time, spatial distribution, and rise rate of the pulse wave), the fluctuation rhythm, force distribution, and waveform morphology of the user's real pulse are dynamically reconstructed in hardware, forming an electronic pulse image that can be perceived by touch. Through high-precision tactile simulation technology, the digitized dynamic pulse image is transformed into a physical pulse that can be touched, perceived, and remotely interacted with. This marks the first time that the objectification, remote accessibility, and standardization of TCM pulse diagnosis have been achieved, providing a verifiable and interactive sensory interface for combining AI with TCM diagnosis.

[0077] In some embodiments, the above-mentioned control of the protrusion time, protrusion position, and protrusion velocity of the target nail-shaped protrusion points based on the dynamic pulse overall simulation 3D model and the distance change data corresponding to each target nail-shaped protrusion point, using the simulation HD vibrator in the pulse dynamic simulator, may specifically include the following steps:

[0078] The dynamic pulse overall fluctuation amplitude and pulse strength in the dynamic pulse overall simulation 3D model are amplified according to the specified magnification ratio, while keeping the overall dynamic pulse fluctuation shape and pulse frequency unchanged, to generate the user's dynamic pulse amplification simulation highlighting model.

[0079] Based on the dynamic pulse amplification simulation protrusion model and the amplified distance change data corresponding to each target nail-shaped protrusion point, the protrusion time, protrusion position and protrusion speed of the target nail-shaped protrusion point are controlled by the simulated HD vibrator in the pulse dynamic simulator.

[0080] By amplifying the amplitude and intensity of a dynamic pulse pattern in a 3D model at a specified ratio (while maintaining the wave pattern and frequency), and using this amplification to drive the HD vibrator in the 3D pulse dynamic simulator to control the motion parameters of the nail-shaped protrusions, the perceptibility and recognizability of weak pulse signals are significantly enhanced. This makes previously imperceptible subtle pulse characteristics (such as thin pulse, soft pulse, and faint pulse) clearly visible at the tactile level, thereby improving the sensitivity, teaching efficiency, and AI diagnostic reliability of traditional Chinese medicine pulse diagnosis. By faithfully amplifying the amplitude and intensity of weak pulses, without distorting their key diagnostic features, this transforms previously imperceptible physiological signals into clear and perceptible physical feedback. For the first time, adjustable sensitivity intelligent pulse diagnosis tactile enhancement has been achieved, greatly improving the accessibility, accuracy, and universality of traditional Chinese medicine pulse recognition.

[0081] In some embodiments, a pressure sensor is provided on the 3D pulse dynamic simulator; the method may further include the following steps: in response to a pulse diagnosis operation acting on the 3D pulse dynamic simulator, detecting the operation pressure data corresponding to the pulse diagnosis operation through the pressure sensor; transmitting the operation pressure data to the airbag controller corresponding to the airbag; and controlling the airbag controller to simulate the operation pressure of the pulse diagnosis operation by using the pressure generated by different airbag inflation volumes according to the operation pressure data.

[0082] In this embodiment, a pressure sensor is integrated into a 3D pulse dynamic simulator, linked to an airbag controller and an adjustable airbag system. The airbag inflation volume is adjusted in real time according to the pulse-taking pressure applied by the physician or user to simulate that pressure in reverse. This constructs a two-way human-computer interactive pulse diagnosis system with realistic tactile feedback and pressure adaptability. This allows TCM practitioners to experience pressure resistance and rebound sensations close to those of a real human wrist during remote or simulated pulse taking, significantly improving the immersion, realism, and diagnostic reliability of the pulse diagnosis simulation. The introduction of a pressure-feedback two-way interactive mechanism into the digital pulse simulation system, through pressure sensing and dynamic airbag control, realistically recreates the mechanical experience of finger palpation in TCM. This enables remote, educational, or AI-assisted pulse diagnosis to move beyond simply observing and listening to the pulse, moving towards true pulse feeling, and greatly enhancing the clinical usability of digital TCM pulse diagnosis.

[0083] In some embodiments, the above-mentioned method of controlling the airbag with an airbag controller to simulate the operating pressure of pulse diagnosis operation by controlling the airbag according to the operating pressure data and generating pressure corresponding to different airbag inflation volumes may specifically include the following steps:

[0084] Based on the operating pressure of the pulse diagnosis procedure, the elastic coefficient of the air bladder, and the contact area between the air bladder and the user's skin, the target inflation volume of the air bladder to be inflated is determined using the following formula, so that the air bladder pressure generated by the target inflation volume can simulate the operating pressure of the pulse diagnosis procedure:

[0085]

[0086] in, Indicates the inflation volume within the target airbag; This indicates the operating pressure during pulse diagnosis. This indicates the contact area between the airbag and the user's skin; This represents the ideal gas constant inside the airbag; This indicates the ambient temperature of the environment in which the airbag is located; This indicates the atmospheric pressure of the environment in which the airbag is located; This indicates the initial inflation volume of the airbag; The elastic coefficient of the airbag is an equivalent stiffness parameter used to characterize the ability of the airbag's membrane material to resist deformation under pressure; the airbag is inflated into the airbag by controlling the inflation of the airbag based on the target inflation volume.

[0087] In this embodiment of the application, the calculation method of the above formula can more efficiently and accurately determine the inflation volume of the target airbag, making the data of the inflation volume of the target airbag more accurate.

[0088] By simultaneously acquiring pulse, EEG, and 3D dynamic behavior models, emotional fluctuations can be captured from both physiological and behavioral dimensions. This multi-source data fusion enhances the reliability of emotion recognition and avoids misjudgment based on a single modality. Furthermore, using a drone equipped with a UWB chip achieves centimeter-level positioning, dynamically acquiring user images from multiple perspectives, and reconstructing a high-fidelity 3D dynamic model to fully reproduce micro-expressions and subtle movements. This high-precision spatial modeling improves the granularity of behavior analysis and overcomes the problems of occlusion and information loss caused by fixed camera perspectives. Moreover, an emotional event is only triggered when at least two modalities of data simultaneously exhibit changes consistent with emotional characteristics within a specified time interval, effectively eliminating random noise or isolated anomalies and achieving temporal consistency verification. With a low false alarm rate, if an emotional event is confirmed, the system immediately associates it with the corresponding environmental audio and video at that moment. Through AI analysis, it identifies specific psychological stimuli, such as specific people, words, or scenes, advancing monitoring from state recognition to causal tracing. This enables precise location of stimuli by tracing back the environmental context. Based on historical stimuli and user response patterns, it dynamically generates targeted monitoring and intervention strategies, improving the system's long-term adaptability and intervention effectiveness. This achieves personalized strategy generation and closed-loop intervention. Therefore, through the aforementioned multimodal perception, spatiotemporal alignment, environmental correlation, and causal inference technologies, it achieves more accurate, in-depth, and interpretable monitoring of users' psychological states, thereby significantly improving the accuracy of mental health monitoring.

[0089] In this embodiment, an AI-based mental health monitoring and intervention system based on the diagnostic principles of Traditional Chinese Medicine is constructed. It integrates multi-source data such as information from "observation, auscultation, inquiry, and palpation," brain function testing data, physiological parameters, behavioral performance, and user self-feedback. It uses a deep learning model to assess mental state, identify potential risks, automatically generate and dynamically adjust personalized intervention plans, and continuously safeguard the user's mental health.

[0090] This invention focuses on the fields of health management, psychological assessment, and traditional Chinese medicine diagnosis. Its core is to integrate the diagnostic ideas of traditional Chinese medicine ("observation, auscultation, inquiry, and palpation") with modern intelligent hardware and artificial intelligence technology to build a mental health monitoring and intervention system. Through real-time collection, deep integration, and intelligent analysis of multi-source data, it can achieve personalized and dynamic psychological state assessment and intervention.

[0091] Regarding system components, cloud server: mobile terminal device: supports Android / iOS system, equipped with ≥7-inch high-definition touch screen (≥1920×1080 resolution), storage ≥64GB, supports Wi-Fi (IEEE 802.11ac), 4G / 5G network.

[0092] The TCM diagnostic instrument for observation, auscultation, inquiry, and palpation includes a multi-source LED array with adjustable color temperature (3000K-6500K) and brightness, a CMOS color sensor with a color difference error of ≤3%, a tongue image scanner with ≥300 dpi and support for multiple angles and focal points, and an optical reflective pulse sensor with a sampling frequency of ≥500Hz and a sensitivity of ≥0.1mmHg.

[0093] Brain function testing equipment: Multi-channel electroencephalography (EEG) acquisition instrument, sampling frequency ≥256Hz, using Ag / AgCl electrodes conforming to the 10-20 system, signal-to-noise ratio ≥30dB. Wearable devices: Smart bracelet / watch, equipped with an optical heart rate sensor with a sampling frequency ≥50Hz and an error of ±2bpm, as well as motion and skin temperature monitoring sensors and an accelerometer with a range of ±16g and a sampling frequency ≥50Hz. Interface layer: Supports BLE 5.0, Wi-Fi 6, and USB-C interfaces to ensure high-speed and stable data transmission. The cloud server establishes communication connections with the TCM "observation, auscultation, inquiry, and palpation" acquisition instrument, brain function testing equipment, and wearable devices through interface layer protocols (supporting BLE 5.0, Wi-Fi 6, and USB-C).

[0094] For the overall process, step 1: Multi-source data reception:

[0095] (1) Receive psychological questionnaire data submitted by users through the mobile terminal front-end interface. This process requires responding to the user's submission operation instructions, verifying the completeness of the questionnaire (such as whether required fields are missing), and ensuring the validity of the basic feedback data. The core content of the questionnaire data directly serves the assessment of psychological state indicators such as "anxiety, depression, and stress". It mainly includes the user's subjective feedback on their own emotions, cognition, and psychological experience, including: psychological state information, subjective physical and mental and behavioral information related to traditional Chinese medicine diagnosis.

[0096] (2) According to the preset sampling frequency, the data receiving mechanism is triggered to capture in real time the color data (captured by CMOS color sensor), tongue image data (≥300 dpi scan image), pulse data (≥500Hz sampling signal of optical reflective pulse sensor), multi-channel EEG data of brain function device (≥256Hz sampling, signal conforming to 10-20 system electrode layout), heart rate (≥50Hz sampling, error ±2bpm), sleep, exercise (accelerometer ±16g range, ≥50Hz sampling) and other physiological parameters of wearable device.

[0097] (3) Transmit the received questionnaire text data, image data collected by the device, physiological electrical signal data, and motion sensor data to the data buffer of the cloud server;

[0098] Further preliminary verification involves checking the data format (e.g., whether the image resolution meets the standard, and whether there are any disconnections or missing signal data) and the consistency of device identification (ensuring that the data is accurately associated with the user's identity). Invalid data is marked and a re-acquisition instruction is triggered (e.g., prompting the user to rescan when the tongue image is blurry).

[0099] Step 2: Data Preprocessing

[0100] (1) Use bandpass filtering (0.5-50Hz) to filter out interference:

[0101] Bandpass filtering logic is applied to physiological electrical signals (pulse, EEG), using a filtering frequency band of 0.5-50Hz. Digital filtering algorithms (such as FIR or IIR filters) are used to filter out noise such as environmental electromagnetic interference, equipment contact noise, and redundant human physiological signals (such as electromyography interference), while retaining the effective signal frequency band.

[0102] (2) Normalization processing: Normalization processing is performed on data of different dimensions and magnitudes. The Z-score or Min-Max normalization algorithm is adaptively selected according to the data characteristics. For physiological parameters with normal distribution (such as mean heart rate), Z-score normalization is used (the data is converted into a standard distribution with mean 0 and standard deviation 1). For data with limited range such as image pixel values ​​and color parameters, Min-Max normalization is used (the data is compressed to the [0,1] interval) to eliminate the influence of data magnitude differences on subsequent model analysis.

[0103] (3) Extracting features such as color, tongue appearance, pulse, and EEG frequency band energy ratio: Analyze color sensor data through image recognition algorithms to extract color features such as the proportion of RGB three color components and color saturation; perform edge detection and texture analysis on tongue scan images to extract tongue coating features such as tongue shape, coating color, and coating texture; perform time domain analysis (extracting pulse wave period, amplitude, and rising slope) and frequency domain analysis on pulse signals to extract pulse features; divide multi-channel EEG data into frequency bands (such as δ, θ, α, β bands) and calculate the energy proportion of each frequency band and the energy ratio between frequency bands; convert textual questionnaire answers into quantitative indicators (such as Likert scale scores), extract keywords from open-ended questions (such as psychological state keywords such as "insomnia" and "irritability") and map them into feature vectors.

[0104] Step 3: Multi-model collaborative reasoning analysis: Call the TCM syndrome differentiation model trained by the deep learning model based on ResNet or EfficientNet architecture, take the preprocessed color, tongue appearance, pulse characteristics, questionnaire quantitative characteristics, and EEG data feature vector as input, and output the TCM syndrome differentiation category (Qi stagnation, blood stasis, Yin deficiency, etc.) and the corresponding probability value (e.g., Qi stagnation probability 0.72, blood stasis probability 0.15), and set the probability threshold to 0.5 (if it is higher than the threshold, it is judged as the corresponding syndrome type).

[0105] The system launches a multi-task LSTM or Transformer model, combined with a psychological state assessment model constructed using random forest or SVM algorithms. It inputs physiological parameters (heart rate variability, sleep structure parameters), behavioral characteristics (exercise frequency, sleep duration), and questionnaire quantitative characteristics collected by wearable devices. The system captures the dynamic change patterns of data through the temporal modeling capabilities of LSTM / Transformer, and enhances the accuracy of indicator assessment by leveraging the classification advantages of random forest / SVM. Finally, it outputs quantitative scores (e.g., anxiety score 65 / 100, depression score 42 / 100) and level determinations (mild, moderate, severe) for psychological state indicators such as anxiety, depression, and stress.

[0106] The risk intervention model employs a multi-task learning architecture, using TCM syndrome differentiation results and psychological state assessment scores as core inputs, while also linking user basic information (age, gender, and past health history). Through the model's shared feature layer, cross-dimensional correlation information is extracted. The risk identification branch outputs the probability of psychological risk occurrence (e.g., severe anxiety risk 0.38) and risk level (low, medium, high). The intervention generation branch calls upon a pre-set intervention strategy knowledge base (including categories such as exercise, diet, and psychological adjustment), and matches and generates personalized intervention measures based on syndrome differentiation (e.g., Yin deficiency corresponding to Yin-nourishing diet suggestions), psychological state level (e.g., moderate stress corresponding to mindfulness training suggestions), and user behavior habits (e.g., exercise preferences).

[0107] The logical steps for generating diagnostic reports and intervention plans are as follows: The results of TCM syndrome differentiation (syndrome name, probability), psychological state assessment indicators (scores and levels of each dimension), risk level, and probability are structurally integrated. Intervention recommendations must be marked with their priority (e.g., "Priority Implementation: 15 minutes of daily mindfulness training," "Supportive Implementation: Tai Chi exercise 3 times per week"). The diagnostic reports and intervention plans are pushed to the user interface in real time via mobile terminal push mechanisms. User feedback: The intervention plan is adjusted based on the assessment results, and the user's status is continuously monitored.

[0108] To address the fixed sampling frequency preset in step 1 (2), a dynamic sampling triggering and parameter adjustment mechanism based on the user's real-time state is added to improve the accuracy and efficiency of data collection. First, the current emotion label (such as "irritable" or "fatigued") actively input by the user is received, while the wearable device monitors physiological warning signals (such as a sudden increase in heart rate ≥100 bpm) in real time. The two types of information are then synchronized to the cloud server in real time as the current emotion state information.

[0109] The cloud server has a built-in sampling strategy library that adjusts the sampling frequency based on the received current emotional state information. Specifically: if "irritability" is detected along with a sudden increase in heart rate, it is determined to be a "high attention state," triggering a sampling frequency upgrade. At this time, the sampling frequency of the brain function detection device increases from 256Hz to 512Hz, the sampling frequency of the pulse sensor of the TCM acquisition instrument increases from 500Hz to 1000Hz, and the tongue scanner triggers "high-definition mode" (resolution increased to 600dpi). If the emotion is "calm" and physiological indicators are stable, it is determined to be a "normal state," maintaining the original sampling frequency or reducing it to energy-saving mode (such as reducing the EEG sampling frequency to 128Hz).

[0110] Data association tagging: The adjusted sampling frequency is novelly bound to the user's current emotional state (such as "high attention - irritability") and transmitted to the data buffer along with the collected data to ensure that the data collection scenario is traceable during subsequent preprocessing and model analysis.

[0111] After the diagnostic report is sent, a feedback window pops up on the mobile device, receiving the user's binary label ("accurate" or "inaccurate") and supplementary description (such as "difficulty falling asleep was not mentioned"). The cloud server associates the feedback labels with the corresponding data (collected data, feature vectors, and inference results) to build an incremental dataset for updating and optimizing the model.

[0112] To strengthen the logical connection between traditional Chinese medicine (TCM) diagnosis and modern psychological assessment, and to enhance the scientific rigor of the core reasoning results, a "verification" step can be added after "the TCM diagnosis model outputs the TCM diagnosis category and corresponding probability value" and "the psychological state assessment model outputs the quantitative score and level of the psychological state index" in step 3, and before "the risk intervention model input." This step includes:

[0113] A pre-constructed "Traditional Chinese Medicine (TCM) Syndrome Type - Psychological State Correlation Validation Rule Base" is established, including positive correlations (such as "Qi stagnation → positive correlation with anxiety / stress" and "Yin deficiency → positive correlation with depressive tendency"), negative exclusions (such as "blood stasis → no direct correlation with mild anxiety" and "phlegm turbidity → weak correlation with simple stress"), and probability matching thresholds (such as when the syndrome type probability is ≥0.6, the corresponding psychological state score must be within ±10 points). The "syndrome type + probability" output by the TCM syndrome differentiation model and the "score + level" output by the psychological state assessment model are input into the validation layer and matched according to the rule base.

[0114] If the results match (e.g., "probability of qi stagnation 0.75 → anxiety score 68 / 100 (moderate)"), output the results directly to the risk intervention model;

[0115] If there is a discrepancy (e.g., "Qi stagnation probability 0.75 → Anxiety score 30 / 100 (mild)"), trigger the following "fine-grained feature re-inference" process (to verify whether the deviation is caused by "incomplete features"):

[0116] The intermediate feature layers of the two models are invoked to extract fine-grained features that were not involved in the initial inference (such as the "rate of change of the slope of the rising edge of the pulse" in the TCM model and the "high-frequency component of heart rate variability" in the psychological model). These features are then re-input into the corresponding models for secondary inference. If the secondary results are consistent, it indicates that the subsequent risk intervention model can directly adopt the result generation scheme.

[0117] If the results are still inconsistent, the results will be corrected by weighting the results with “device data feature weight (0.6) + questionnaire data feature weight (0.4)” (e.g., if the Qi stagnation syndrome is clear, the anxiety score will be adjusted to a reasonable range). Objective equipment data (TCM data collection instrument, wearable device) will be given priority, while subjective feedback will be taken into account.

[0118] In this embodiment, multi-channel and multi-modal data fusion is achieved: improving the comprehensiveness and accuracy of psychological state assessment; integration with traditional Chinese medicine diagnostic thinking to enhance the systematicness and personalization of the system; automated personalized intervention: dynamically adjusting intervention measures based on risk assessment results to improve intervention effectiveness; and continuous monitoring capability: realizing dynamic management and early warning of mental health.

[0119] Figure 3 A schematic diagram of an AI-powered mental health monitoring device based on Traditional Chinese Medicine (TCM) data is provided. This device can be applied to a monitoring system, which is connected to a pulse sensor, a camera, and an electroencephalogram (EEG) scanner. The camera is mounted on a drone and is equipped with an ultra-wideband chip. Figure 3 As shown, the AI-powered mental health monitoring device 300 based on Traditional Chinese Medicine data includes:

[0120] The acquisition module 301 is used to acquire the user's pulse data through the pulse sensor and to acquire the user's brainwave data through the brainwave acquisition device.

[0121] The generation module 302 is used to control the UAV to fly to multiple different locations around the user based on the real-time positioning of the ultra-wideband chip, determine multiple different relative positional relationships between the multiple locations and the user using the ultra-wideband chip, acquire dynamic images of the user and surrounding environment videos of the corresponding time period through the camera at each location, and generate a 3D dynamic model of the user based on the multiple relative positional relationships and the corresponding acquired dynamic images; the dynamic images include facial change images, body posture change images, and action execution images of the user;

[0122] The analysis module 303 is used to respond to at least two real-time change data in the 3D dynamic model, the pulse data, and the EEG data that conform to specified psychological and emotional fluctuation phenomena data, and to determine whether the interval between the occurrence time of the change corresponding to the at least two real-time change data is less than a specified time interval. If it is less than the specified time interval, the module determines the surrounding environment voice content and surrounding environment image content corresponding to the occurrence time of the change from the surrounding environment video, and analyzes the user's psychological stimulus points through the AI ​​system based on the surrounding environment voice content and the surrounding environment image content, and generates a psychological health monitoring strategy for the user based on several of the psychological stimulus points.

[0123] The AI ​​mental health monitoring device based on traditional Chinese medicine data provided in this application embodiment has the same technical features as the AI ​​mental health monitoring method based on traditional Chinese medicine data provided in the above embodiment, so it can also solve the same technical problems and achieve the same technical effects.

[0124] An electronic device provided in this application embodiment, such as Figure 4 As shown, the electronic device 400 includes a processor 402 and a memory 401. The memory stores a computer program that can run on the processor. When the processor executes the computer program, it implements the steps of the method provided in the above embodiments.

[0125] See Figure 4 The electronic device also includes a bus 403 and a communication interface 404. The processor 402, the communication interface 404 and the memory 401 are connected via the bus 403. The processor 402 is used to execute executable modules, such as computer programs, stored in the memory 401.

[0126] The memory 401 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 404 (which can be wired or wireless), such as the Internet, wide area network, local area network, or metropolitan area network.

[0127] Bus 403 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0128] The memory 401 is used to store programs. After receiving an execution instruction, the processor 402 executes the program. The method executed by the apparatus defined by the process disclosed in any of the preceding embodiments of this application can be applied to the processor 402 or implemented by the processor 402.

[0129] Processor 402 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 402 or by instructions in software form. The processor 402 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 401, and processor 402 reads the information from memory 401 and, in conjunction with its hardware, completes the steps of the above method.

[0130] Corresponding to the above-described AI-based mental health monitoring method based on traditional Chinese medicine data, this application embodiment also provides a computer-readable storage medium storing computer-executable instructions. When the computer-executable instructions are invoked and executed by a processor, the computer-executable instructions cause the processor to perform the steps of the above-described AI-based mental health monitoring method based on traditional Chinese medicine data.

[0131] The AI-based mental health monitoring device based on traditional Chinese medicine data provided in this application embodiment can be specific hardware on a device or software or firmware installed on the device. The implementation principle and technical effects of the device provided in this application embodiment are the same as those in the foregoing method embodiments. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the foregoing method embodiments. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can all be referred to the corresponding processes in the above method embodiments, and will not be repeated here.

[0132] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0133] For example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0134] 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.

[0135] In addition, the functional units in the embodiments provided in 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.

[0136] 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 AI ​​mental health monitoring method based on traditional Chinese medicine data 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.

[0137] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0138] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application. All should be covered within the protection scope of this application. Therefore, the protection scope of this application should be determined by the protection scope of the claims.

Claims

1. An AI-based mental health monitoring method based on traditional Chinese medicine data, characterized in that, The method is applied to a monitoring system, which is correspondingly connected to a pulse sensor, a camera, and an electroencephalogram (EEG) collector. The camera is mounted on a drone and is equipped with an ultra-wideband chip. The pulse sensor collects the user's pulse data, and the brainwave collector collects the user's brainwave data. The UAV is controlled to fly to multiple different locations around the user based on the real-time positioning of the ultra-wideband chip. The UAV chip is used to determine multiple different relative positional relationships between the multiple locations and the user. Based on each location, the camera acquires dynamic images of the user and surrounding environment videos for the corresponding time period of the dynamic images. Based on the multiple relative positional relationships and the corresponding acquired dynamic images, a 3D dynamic model of the user is generated. The dynamic images include images of facial changes, body posture changes, and action execution. In response to at least two real-time change data in the 3D dynamic model, the pulse data, and the EEG data conforming to specified psychological and emotional fluctuation phenomena, it is determined whether the interval between the occurrence times of the changes corresponding to the at least two real-time change data is less than a specified time interval. If it is less than the specified time interval, the surrounding environment voice content and surrounding environment image content corresponding to the occurrence time of the change are determined from the surrounding environment video. Based on the surrounding environment voice content and surrounding environment image content, the user's psychological stimulus points are analyzed by the AI ​​system, and a psychological health monitoring strategy for the user is generated based on several of the psychological stimulus points. The step of generating a mental health monitoring strategy for a user based on a plurality of psychological stimulus point objects includes: generating a mental health monitoring strategy for a user based on a plurality of psychological stimulus point objects and a target dynamic pulse overall simulation 3D model corresponding to the occurrence time of each psychological stimulus point object; the monitoring system is also connected to a tongue image scanner, a respiratory flow sensor and a taste sensor. The method of generating a mental health monitoring strategy for a user based on a set of psychological stimulation points and a target dynamic pulse overall simulation 3D model corresponding to the occurrence time of each psychological stimulation point includes: collecting the user's tongue image data through the tongue image scanner; converting the user's inhaled and exhaled gas flow rate into electrical signals through the respiratory flow sensor and detecting the user's unit time ventilation, breath flow rate parameter, tidal volume, breath temperature, and oxygen and carbon dioxide concentration based on the electrical signals; collecting the user's body odor data through the odor sensor; analyzing the user's visceral pathological data through the AI ​​system based on the dynamic pulse overall simulation 3D model, the 3D dynamic model, the EEG data, the tongue image data, the unit time ventilation, the breath flow rate parameter, the tidal volume, the breath temperature, the body odor data, and the oxygen and carbon dioxide concentration; and generating a mental health monitoring plan for the user based on the visceral pathological data, the set of psychological stimulation points, and the target dynamic pulse overall simulation 3D model corresponding to the occurrence time of each psychological stimulation point.

2. The method according to claim 1, characterized in that, The pulse sensor is mounted on the overall pulse detection wristband. An ultrasonic radar measuring device and an airbag are located on the inner side of the wristband. The ultrasonic radar measuring device is positioned inside the airbag near the wristband, with its measurement direction facing inwards. The thickness of the airbag's film is less than a specified thickness. The generation of a user's mental health monitoring strategy based on several psychological stimulus points includes: The pulse sensor uses piezoelectricity to detect the user's pulse frequency and pulse strength. When the degree of deformation caused by the inflation volume in the airbag meets the specified degree of deformation and the deformation stops, the overall fluctuation amplitude and overall fluctuation pattern of the user's dynamic pulse are determined based on multiple distance change data measured by the ultrasonic radar measuring device; wherein, the multiple distance change data includes the distance changes between multiple skin position points within the overall skin area corresponding to the user's wrist pulse range as measured by the ultrasonic radar measuring device. Based on the overall dynamic pulse wave pattern, the overall dynamic pulse wave amplitude, the pulse frequency, and the pulse strength, a 3D simulation model of the user's overall dynamic pulse wave is generated. A mental health monitoring strategy for users is generated based on a 3D simulation model of the overall dynamic pulse corresponding to the occurrence time of each of the aforementioned psychological stimulus points.

3. The method according to claim 2, characterized in that, After generating a 3D simulation model of the user's overall dynamic pulse based on the overall fluctuation pattern of the dynamic pulse, the overall fluctuation amplitude of the dynamic pulse, the pulse frequency, and the pulse strength, the method further includes: The dynamic pulse pattern overall simulation 3D model, the multiple skin location points, and the distance change data corresponding to each skin location point are transmitted to the 3D pulse pattern dynamic simulator; wherein, the surface of the 3D pulse pattern dynamic simulator is provided with multiple nail-shaped protrusions. Determine the target nail-shaped protrusion point from the plurality of nail-shaped protrusion points that corresponds to the plurality of skin location points; Based on the overall dynamic pulse simulation 3D model and the distance change data corresponding to each of the target nail-shaped protrusion points, the simulation HD vibrator in the pulse dynamic simulator is used to control the protrusion time, protrusion position, and protrusion speed of the target nail-shaped protrusion points. The overall dynamic pulse fluctuation of the user is simulated by the protrusion time, protrusion position, and protrusion speed of the target nail-shaped protrusion points, so as to perform TCM pulse diagnosis based on the overall dynamic pulse fluctuation.

4. The method according to claim 3, characterized in that, The method of controlling the protrusion time, protrusion position, and protrusion speed of the target nail-shaped protrusions using the simulated HD vibrator in the pulse dynamic simulator, based on the overall 3D simulation model of the dynamic pulse and the distance change data corresponding to each of the target nail-shaped protrusions, includes: According to the specified magnification ratio, the overall fluctuation amplitude and pulse strength of the dynamic pulse in the overall simulation 3D model of the dynamic pulse are magnified while keeping the overall fluctuation shape and pulse frequency of the dynamic pulse unchanged, so as to generate the user's dynamic pulse magnification simulation highlighting model. Based on the dynamic pulse amplification simulation protrusion model and the amplified distance change data corresponding to each of the target nail-shaped protrusion points, the protrusion time, protrusion position, and protrusion speed of the target nail-shaped protrusion points are controlled by the simulated HD vibrator in the pulse dynamic simulator.

5. The method according to claim 3, characterized in that, The 3D pulse dynamic simulator is equipped with a pressure sensor; the method further includes: In response to the pulse diagnosis operation applied to the 3D pulse dynamic simulator, the pressure sensor detects the operation pressure data corresponding to the pulse diagnosis operation. The operating pressure data is transmitted to the airbag controller corresponding to the airbag. The airbag controller controls the airbag to simulate the operating pressure of the pulse diagnosis operation by adjusting the pressure generated by different airbag inflation volumes according to the operating pressure data.

6. The method according to claim 5, characterized in that, The step of controlling the airbag controller to simulate the operating pressure of the pulse diagnosis operation by adjusting the pressure generated by different airbag inflation volumes according to the operating pressure data includes: Based on the operating pressure of the pulse diagnosis operation, the elastic coefficient of the airbag, and the contact area between the airbag and the user's skin, the target inflation volume of the airbag to be inflated is determined using the following formula, so that the airbag pressure generated by the target inflation volume can simulate the operating pressure of the pulse diagnosis operation: in, This indicates the inflation volume within the target airbag; This indicates the operating pressure of the pulse diagnosis operation; This indicates the contact area between the airbag and the user's skin; This represents the ideal gas constant within the airbag; This indicates the ambient temperature of the environment in which the airbag is located; This indicates the atmospheric pressure of the environment in which the airbag is located; This indicates the initial inflation volume within the airbag; The elastic coefficient of the airbag is an equivalent stiffness parameter used to characterize the ability of the thin film material of the airbag to resist deformation under pressure. Based on the target air volume, the airbag controller controls the inflation of the airbag.

7. The method according to claim 1, characterized in that, The EEG acquisition device is a multi-channel EEG acquisition device; the specified psychological and emotional fluctuation data includes at least one of the following: Data on changes in body posture, movements, facial expressions, pulse, and various types of electroencephalograms when psychological and emotional fluctuations occur.

8. An AI-powered mental health monitoring device based on Traditional Chinese Medicine (TCM) data, characterized in that, The device is applied to a monitoring system, which is correspondingly connected to a pulse sensor, a camera, and an electroencephalogram (EEG) collector. The camera is mounted on a drone and is equipped with an ultra-wideband chip. The device includes: The acquisition module is used to acquire the user's pulse data through the pulse sensor and to acquire the user's brainwave data through the brainwave acquisition device. The generation module is used to control the drone to fly to multiple different locations around the user based on the real-time positioning of the ultra-wideband chip, determine multiple different relative positional relationships between the multiple locations and the user using the ultra-wideband chip, acquire dynamic images of the user and surrounding environment videos of the corresponding time period through the camera at each location, and generate a 3D dynamic model of the user based on the multiple relative positional relationships and the corresponding acquired dynamic images; the dynamic images include images of facial changes, body posture changes, and action execution. The analysis module is used to respond to at least two real-time change data in the 3D dynamic model, the pulse data, and the EEG data that conform to specified psychological and emotional fluctuation phenomena data, and to determine whether the interval between the occurrence time of the change corresponding to the at least two real-time change data is less than a specified time interval. If it is less than the specified time interval, the module determines the surrounding environment voice content and surrounding environment image content corresponding to the occurrence time of the change from the surrounding environment video, and analyzes the user's psychological stimulus points based on the surrounding environment voice content and surrounding environment image content through the AI ​​system, and generates a psychological health monitoring strategy for the user based on several of the psychological stimulus points. The analysis module further includes: generating a psychological health monitoring strategy for the user based on a 3D simulation model of the overall target dynamic pulse corresponding to the occurrence time of each of the psychological stimulus points; the monitoring system is also connected to a tongue scanner, a respiratory flow sensor and a taste sensor. The analysis module specifically includes: collecting the user's tongue image data through the tongue image scanner; converting the user's inhaled and exhaled gas flow rate into electrical signals through the respiratory flow sensor and detecting the user's ventilation per unit time, breath velocity parameters, tidal volume, breath temperature, and oxygen and carbon dioxide concentrations based on the electrical signals; collecting the user's body odor data through the odor sensor; analyzing the user's visceral pathological data through the AI ​​system based on the dynamic pulse overall simulation 3D model, the 3D dynamic model, the EEG data, the tongue image data, the ventilation per unit time, the breath velocity parameters, the tidal volume, the breath temperature, the body odor data, and the oxygen and carbon dioxide concentrations; and generating a mental health monitoring plan for the user based on the visceral pathological data, several psychological stimulation points, and the target dynamic pulse overall simulation 3D model corresponding to the occurrence time of each psychological stimulation point.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method described in any one of claims 1 to 7.

Images