A non-suicidal self-injury intervention system based on an electric stimulation bracelet

Abstract

The application provides a non-suicidal self-injury behavior intervention system based on an electric stimulation bracelet, which comprises: a physiological characteristic data set is generated by conditioning, demodulating and analog-digital converting an excitation current and an induced current signal in a collection process; a programmable electric stimulation chip is used for multi-parameter fusion calculation on the physiological characteristic data set to generate an emotional stress value, and the emotional stress value is compared with a preset emotional stress threshold value to obtain an intervention trigger signal; when the intervention trigger signal is in an effective state, the programmable electric stimulation chip generates an electric stimulation control signal, and a dry electrode in contact with a human body is used for outputting an electric stimulation current with preset parameters to the human body to execute active intervention on non-suicidal self-injury behavior. The application can greatly improve intervention efficiency.

Description

Technical Field

[0001] This invention relates to the field of electrical stimulation technology, and in particular to a non-suicidal self-harm intervention system based on an electrical stimulation bracelet. Background Technology

[0002] Non-suicidal self-injury (NSSI) is the intentional and direct destruction of one's own body without suicidal intent. Common forms include cuts, burns, and impacts. Studies show that the lifetime prevalence of NSSI is increasing among adolescents and young adults. NSSI not only causes physical damage and permanent scarring but is also a significant predictor of future suicidal behavior, significantly increasing the risk of suicide, making clinical intervention urgently needed.

[0003] Currently, clinical neuromodulation intervention devices for NSSI have significant limitations in application scenarios. Existing TENS-type intervention devices are mostly large, fixed devices installed in hospitals. They are bulky, inconvenient to carry, and can only be used for timed interventions within medical institutions, failing to cover patients' daily lives outside the hospital. Furthermore, NSSI self-harm behaviors are often instantaneous and impulsive, and the lack of immediate intervention methods after patients leave the medical setting is a core reason for NSSI relapse. Currently, there is no portable, emergency-specific stimulation solution adapted to a wristband format.

[0004] Furthermore, existing intervention models are all timed and passive, lacking immediate intervention devices that support patients to actively trigger impulses. They cannot quickly implement interventions during the golden window period when self-harm impulses occur, making it difficult to prevent self-harm behavior from the source. They can only achieve post-event repair, and the timeliness of intervention has significant shortcomings, failing to meet the actual needs of NSSI clinical emergency intervention. Summary of the Invention

[0005] Based on this, the purpose of this invention is to provide a non-suicidal self-harm intervention system based on an electrical stimulation bracelet, in order to solve the problem that the intervention modes of the existing technology are all timed and passive interventions, lacking a method to support the patient's active triggering of impulses for immediate intervention, resulting in the inability to quickly implement intervention during the golden window period when self-harm impulses occur, which has significant shortcomings.

[0006] The first aspect of the present invention proposes: A non-suicidal self-harm intervention method based on an electrical stimulation wristband, wherein the method includes: The wristband collects physiological signals from the wrist via its built-in AFE, and processes the excitation current and induced current signals during the collection process through conditioning, demodulation, and analog-to-digital conversion to generate a physiological feature dataset. The physiological feature dataset is then transmitted to the programmable electrical stimulation chip via SPI communication. The physiological feature dataset is fused and calculated using a programmable electrical stimulation chip to generate an emotional stress value. The emotional stress value is then compared with a preset emotional stress threshold to obtain an intervention trigger signal. When the intervention trigger signal is in an effective state, the programmable electrical stimulation chip generates an electrical stimulation control signal and outputs an electrical stimulation current with preset parameters to the human body through the dry electrodes in contact with the human body in order to perform active intervention for non-suicidal self-harm behavior. The wristband displays its operating status data on a screen and collects user stop commands via a button switch to control the start and stop of the electrical stimulation current output.

[0007] The beneficial effects of this invention are as follows: This solution overcomes the shortcomings of existing technologies where timed, passive interventions cannot respond quickly during the golden window period when self-harm impulses arise. It collects physiological signals from the wrist via an AFE built into the wristband, generates a physiological feature dataset through signal conditioning, demodulation, and analog-to-digital conversion, and then uses a programmable electrical stimulation chip to perform multi-parameter fusion calculations to obtain emotional stress values. By comparing these values ​​with preset thresholds, the intervention trigger signal is automatically determined. When a self-harm impulse suddenly occurs, an electrical stimulation control signal is generated immediately, and a preset electrical stimulation current is output to the body via dry electrodes, achieving proactive and immediate intervention for non-suicidal self-harm behaviors. Simultaneously, the working status is displayed on a screen, and a button switch collects stop commands, allowing users to actively control the start and stop of the intervention, effectively improving the timeliness, accuracy, and safety of the intervention.

[0008] Furthermore, the steps of acquiring physiological signals from the wrist via the built-in AFE of the wristband, and conditioning, demodulating, and performing analog-to-digital conversion on the excitation current and induced current signals during the acquisition process include: Heart rate light-sensing signals are collected and converted into current signals, which are then conditioned by AFE and converted into digital signals to calculate heart rate and heart rate variability indices. A weak AC excitation current with preset parameters is injected into the skin by an AC constant current source, the voltage drop signal across the skin is measured, and the voltage drop signal is demodulated and calculated by the I / Q demodulation circuit inside the AFE to obtain the skin resistance value. The resistance signal of the NTC is collected and calculated using a voltage divider circuit and the B-value formula to obtain the skin temperature value. Heart rate, heart rate variability, skin resistance, and skin temperature data are serialized to generate a physiological feature dataset, which is then transmitted to a programmable electrical stimulation chip via SPI communication.

[0009] Furthermore, the step of performing multi-parameter fusion calculations on the physiological feature dataset using the programmable electrical stimulation chip to generate an emotional stress value includes: Four types of feature parameters—heart rate, heart rate variability, skin resistance, and skin temperature—are extracted from the physiological feature dataset, and the stress contribution value of each type of feature parameter is calculated based on a preset emotional stress assessment model. For feature parameters whose pressure contribution value is higher than the preset weight threshold, a weighted fusion calculation is performed to generate core pressure feature values; For feature parameters whose pressure contribution value is lower than the preset weight threshold, a correlation correction link with the core pressure feature value is established through a feature calibration algorithm. Based on the core stress characteristic value, and combined with the correction amount of the associated correction link, the corresponding emotional stress value is generated.

[0010] Furthermore, the step of generating an electrical stimulation control signal by the programmable electrical stimulation chip and outputting an electrical stimulation current of preset parameters to the human body through dry electrodes in contact with the human body when the intervention trigger signal is in an effective state includes: The intervention trigger signals are labeled with their types to distinguish between single emergency triggers, continuous high-pressure triggers, and abnormal triggers, and the duration and frequency of each type of trigger are statistically analyzed. Based on the type and duration of the trigger, the electrical stimulation output mode is determined, and the key parameters of frequency, pulse width, and current intensity contained in the electrical stimulation output mode are extracted. Several output control sub-items are then divided according to the priority of the key parameters. Based on the hierarchical relationship between the electrical stimulation output mode and several output control sub-items, an electrical stimulation control sequence is constructed, and the output channel ID of the dry electrode is marked at each node of the electrical stimulation control sequence to generate the corresponding electrical stimulation control signal.

[0011] Furthermore, the step of marking the output channel ID of the dry electrode at each node of the electrical stimulation control sequence to generate the corresponding electrical stimulation control signal includes: Each node of the electrical stimulation control timing sequence is hierarchically encoded to form a three-dimensional control code that includes stimulation stage identifiers, timing node identifiers, and output channel identifiers. The output channel IDs of the dry electrode are grouped and clustered according to their output polarity, and a mapping relationship between each output channel ID and the three-dimensional control code is constructed. According to the mapping relationship, the output channel ID is written into the control field of each node of the electrical stimulation control sequence, and the current intensity value and duration weight value of the node are marked next to each output channel ID to generate the corresponding electrical stimulation control signal.

[0012] Furthermore, the steps of displaying the wristband's working status data on the screen and collecting the user's stop command via the button switch include: The physiological feature dataset, emotional stress value, and electrical stimulation state data are classified to generate a state dataset containing working mode, physiological indicators, stimulation state, and device power. The input signal of the button switch is analyzed by edge detection to generate a trigger instruction set. The state dataset is hierarchically matched with the display partitions of the display screen to determine the display priority of each display partition. The trigger instruction set is compared with the preset control instructions by attributes, and valid instructions with a matching degree exceeding the preset matching threshold are retained. High-priority status data is formatted for display to generate display frame data, and valid instructions are parsed to generate a set of control instructions. The display frame data and the set of control instructions are then integrated to generate interactive control signals.

[0013] Furthermore, the step of integrating the display frame data and the control command set to generate an interactive control signal includes: Add display timing tags to the display frame data and add appropriate instruction priority tags to the control instruction set; An interactive control index table is created based on the display timing tag and the instruction priority tag, and a state association matrix is ​​generated based on the display frame data and the control instruction set. The control parameters are extracted from the state association matrix according to the interactive control index table, and the control parameters are set as interactive control signals.

[0014] The second aspect of the present invention proposes: A non-suicidal self-harm intervention system based on an electrical stimulation wristband, wherein the system comprises: The processing module is used to collect physiological signals from the wrist via the built-in AFE of the wristband, and to condition, demodulate and convert the excitation current and induced current signals during the collection process to generate a physiological feature dataset. The physiological feature dataset is then transmitted to the programmable electrical stimulation chip via SPI communication. The calculation module is used to perform multi-parameter fusion calculation on the physiological feature dataset through the programmable electrical stimulation chip to generate an emotional stress value, and compare the emotional stress value with a preset emotional stress threshold to obtain an intervention trigger signal; The intervention module is used to generate an electrical stimulation control signal when the intervention trigger signal is in an effective state, and output an electrical stimulation current with preset parameters to the human body through dry electrodes in contact with the human body in order to perform active intervention for non-suicidal self-harm behavior. The control module is used to display the wristband's working status data on the screen and to collect user stop commands via button switches, thereby controlling the start and stop of the electrical stimulation current output.

[0015] Furthermore, the steps of acquiring physiological signals from the wrist via the built-in AFE of the wristband, and conditioning, demodulating, and performing analog-to-digital conversion on the excitation current and induced current signals during the acquisition process include: Heart rate light-sensing signals are collected and converted into current signals, which are then conditioned by AFE and converted into digital signals to calculate heart rate and heart rate variability indices. A weak AC excitation current with preset parameters is injected into the skin by an AC constant current source, the voltage drop signal across the skin is measured, and the voltage drop signal is demodulated and calculated by the I / Q demodulation circuit inside the AFE to obtain the skin resistance value. The resistance signal of the NTC is collected and calculated using a voltage divider circuit and the B-value formula to obtain the skin temperature value. Heart rate, heart rate variability, skin resistance, and skin temperature data are serialized to generate a physiological feature dataset, which is then transmitted to a programmable electrical stimulation chip via SPI communication.

[0016] Furthermore, the step of performing multi-parameter fusion calculations on the physiological feature dataset using the programmable electrical stimulation chip to generate an emotional stress value includes: Four types of feature parameters—heart rate, heart rate variability, skin resistance, and skin temperature—are extracted from the physiological feature dataset, and the stress contribution value of each type of feature parameter is calculated based on a preset emotional stress assessment model. For feature parameters whose pressure contribution value is higher than the preset weight threshold, a weighted fusion calculation is performed to generate core pressure feature values; For feature parameters whose pressure contribution value is lower than the preset weight threshold, a correlation correction link with the core pressure feature value is established through a feature calibration algorithm. Based on the core stress characteristic value, and combined with the correction amount of the associated correction link, the corresponding emotional stress value is generated.

[0017] Furthermore, the step of generating an electrical stimulation control signal by the programmable electrical stimulation chip and outputting an electrical stimulation current of preset parameters to the human body through dry electrodes in contact with the human body when the intervention trigger signal is in an effective state includes: The intervention trigger signals are labeled with their types to distinguish between single emergency triggers, continuous high-pressure triggers, and abnormal triggers, and the duration and frequency of each type of trigger are statistically analyzed. Based on the type and duration of the trigger, the electrical stimulation output mode is determined, and the key parameters of frequency, pulse width, and current intensity contained in the electrical stimulation output mode are extracted. Several output control sub-items are then divided according to the priority of the key parameters. Based on the hierarchical relationship between the electrical stimulation output mode and several output control sub-items, an electrical stimulation control sequence is constructed, and the output channel ID of the dry electrode is marked at each node of the electrical stimulation control sequence to generate the corresponding electrical stimulation control signal.

[0018] Furthermore, the step of marking the output channel ID of the dry electrode at each node of the electrical stimulation control sequence to generate the corresponding electrical stimulation control signal includes: Each node of the electrical stimulation control timing sequence is hierarchically encoded to form a three-dimensional control code that includes stimulation stage identifiers, timing node identifiers, and output channel identifiers. The output channel IDs of the dry electrode are grouped and clustered according to their output polarity, and a mapping relationship between each output channel ID and the three-dimensional control code is constructed. According to the mapping relationship, the output channel ID is written into the control field of each node of the electrical stimulation control sequence, and the current intensity value and duration weight value of the node are marked next to each output channel ID to generate the corresponding electrical stimulation control signal.

[0019] Furthermore, the steps of displaying the wristband's working status data on the screen and collecting the user's stop command via the button switch include: The physiological feature dataset, emotional stress value, and electrical stimulation state data are classified to generate a state dataset containing working mode, physiological indicators, stimulation state, and device power. The input signal of the button switch is analyzed by edge detection to generate a trigger instruction set. The state dataset is hierarchically matched with the display partitions of the display screen to determine the display priority of each display partition. The trigger instruction set is compared with the preset control instructions by attributes, and valid instructions with a matching degree exceeding the preset matching threshold are retained. High-priority status data is formatted for display to generate display frame data, and valid instructions are parsed to generate a set of control instructions. The display frame data and the set of control instructions are then integrated to generate interactive control signals.

[0020] Furthermore, the step of integrating the display frame data and the control command set to generate an interactive control signal includes: Add display timing tags to the display frame data and add appropriate instruction priority tags to the control instruction set; An interactive control index table is created based on the display timing tag and the instruction priority tag, and a state association matrix is ​​generated based on the display frame data and the control instruction set. The control parameters are extracted from the state association matrix according to the interactive control index table, and the control parameters are set as interactive control signals.

[0021] The third aspect of the present invention proposes: An electrical stimulation bracelet includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the functions of a non-suicidal self-harm intervention system based on the electrical stimulation bracelet as described above.

[0022] The fourth aspect of the present invention proposes: A computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the functions of the non-suicidal self-harm intervention system based on an electrically stimulated wristband as described above.

[0023] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0024] Figure 1 A flowchart of a non-suicidal self-harm behavior intervention method based on an electrical stimulation bracelet provided in the first embodiment of the present invention; Figure 2 A flowchart of a non-suicidal self-harm intervention method based on an electrical stimulation bracelet provided in the second embodiment of the present invention; Figure 3 This is a structural block diagram of a non-suicidal self-harm intervention system based on an electrical stimulation wristband, provided in the third embodiment of the present invention.

[0025] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0026] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0027] It should be noted that when a component is said to be "fixed to" another component, it can be directly on the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0028] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0029] Please see Figure 1 The image shows a non-suicidal self-harm intervention method based on an electrical stimulation bracelet provided in the first embodiment of the present invention. This non-suicidal self-harm intervention method based on an electrical stimulation bracelet can generate an electrical stimulation control signal in real time when a self-harm impulse suddenly occurs, and output an electrical stimulation current with preset parameters to the human body through dry electrodes to achieve active and real-time intervention of non-suicidal self-harm behavior; at the same time, the working status is displayed on the screen and the button switch collects the stop command, supporting the user to actively control the start and stop of the intervention, effectively improving the timeliness, accuracy and safety of the intervention.

[0030] The electrical stimulation bracelet provided in this embodiment includes a multimodal health sensor AFE, a programmable electrical stimulation chip, dry electrodes, a display screen, an LED (light-emitting diode), a PD (photodiode), an NTC (thermistor), and a button switch.

[0031] Specifically, this embodiment provides: A non-suicidal self-harm intervention method based on an electrical stimulation wristband, wherein the method includes: Step S10: The physiological signals at the wrist are collected by the built-in AFE of the wristband. The excitation current and induced current signals during the collection process are conditioned, demodulated and analog-to-digital converted to generate a physiological feature dataset. The physiological feature dataset is then transmitted to the programmable electrical stimulation chip via SPI communication. It should be noted that the wristband's built-in AFE collects physiological signals from the wrist, and the excitation and induced current signals during the collection process are conditioned, demodulated, and converted from analog to digital to generate a physiological feature dataset. This dataset is then transmitted to a programmable electrical stimulation chip via SPI communication. This step is the perceptual foundation of the entire method, capturing the peripheral physiological representations of emotional stress through multimodal physiological signals, providing objective data input for subsequent stress assessment.

[0032] The wristband's sensors and electrodes are all positioned on the inner side of the wrist, corresponding to the distribution areas of the median and ulnar nerves. This ensures stable contact between the PPG photoplethysmography (PPG) sensor, GSR excitation electrode, and NTC thermistor with the skin. The PPG sensor uses a green LED, corresponding to the highly vascularized inner side of the wrist, reducing motion artifacts. The GSR employs a four-electrode method: two excitation electrodes inject current, and two measurement electrodes collect voltage, improving measurement accuracy. The NTC thermistor is embedded in the electrode base, closely adhering to the skin to ensure accurate temperature measurement.

[0033] Step S20: The physiological feature dataset is fused and calculated using a programmable electrical stimulation chip to generate an emotional stress value. The emotional stress value is then compared with a preset emotional stress threshold to obtain an intervention trigger signal. It is worth noting that a programmable electrical stimulation chip is used to perform multi-parameter fusion calculations on the physiological feature dataset to generate an emotional stress value. This emotional stress value is then compared with a preset emotional stress threshold to obtain an intervention trigger signal. This step is the core of the method's decision-making process, overcoming the limitations of a single physiological indicator. Multi-parameter cross-validation improves the accuracy of emotional stress assessment, avoiding false triggers caused by interference from movement, temperature, and individual differences. Simultaneously, it enables the grading and quantification of stress, providing a basis for subsequent graded interventions.

[0034] Step S30: When the intervention trigger signal is in an effective state, the programmable electrical stimulation chip generates an electrical stimulation control signal and outputs an electrical stimulation current with preset parameters to the human body through the dry electrodes in contact with the human body in order to perform active intervention for non-suicidal self-harm behavior. It should be noted that when the intervention trigger signal is valid, the programmable electrical stimulation chip generates an electrical stimulation control signal, which outputs a preset electrical stimulation current to the body through dry electrodes in contact with the body, performing active intervention against non-suicidal self-harm behavior. This step is the core of the method, outputting electrical stimulation in a corresponding mode according to the pressure level and trigger type, regulating nerve excitability through transcutaneous electrical stimulation, and interrupting the chain of self-harm impulses. The electrical stimulation adopts a constant current output design to ensure current stability when skin impedance changes, and multiple safety thresholds are set to avoid current overload and skin damage.

[0035] Step S40: Display the wristband's working status data on the screen, and collect the user's stop command through the button switch to control the start and stop of the electrical stimulation current output.

[0036] It's worth noting that the wristband's operating status data is displayed on a screen, and user stop commands are collected via a button switch, corresponding to the start and stop of the electrical stimulation current output. This step ensures the method's safety and interactivity. The status display allows users to understand their current physiological indicators and stimulation status, while the physical buttons enable users to autonomously control the intervention. It also supports mode switching, manual triggering, and other functions, balancing intervention effectiveness with user autonomy and improving compliance.

[0037] Please see Figure 2 The second embodiment of the present invention provides: Furthermore, the steps of acquiring physiological signals from the wrist via the built-in AFE of the wristband, and conditioning, demodulating, and performing analog-to-digital conversion on the excitation current and induced current signals during the acquisition process include: Step S101: Acquire the heart rate photosensitive signal and convert it into an electrical signal, then convert it into a digital signal through AFE conditioning to calculate the heart rate and heart rate variability index. Step S102: A weak AC excitation current with preset parameters is injected into the skin through an AC constant current source, the voltage drop signal across the skin is measured, and the voltage drop signal is demodulated and calculated through the I / Q demodulation circuit inside the AFE to obtain the skin resistance value. Step S103: Acquire the resistance signal of the NTC, calculate the skin temperature value using a voltage divider circuit and the B-value formula; serialize the heart rate, heart rate variability, skin resistance, and skin temperature data to generate a physiological feature dataset, and transmit it to the programmable electrical stimulation chip via SPI communication.

[0038] It is important to note that heart rate and heart rate variability (HRV) are core indicators reflecting the sympathetic / parasympathetic balance of the autonomic nervous system. When emotional stress increases or self-inflicted impulses arise, the sympathetic nervous system is excited, heart rate increases, RR interval variability decreases, especially the high-frequency (HF) component, and the low-frequency (LF) / high-frequency (HF) ratio increases, making it a sensitive physiological marker of emotional arousal. Compared to a single heart rate indicator, HRV is more specific to emotional stress and is less affected by exercise and body temperature.

[0039] Employing PPG (Photoplethysmography) technology, the wristband incorporates a 525nm green LED and a high-sensitivity photodiode. The LED is driven by a 10mA pulsed current to reduce power consumption. The LED emits green light that illuminates the dermis, and the blood volume changes periodically with the heartbeat, causing a synchronous change in the intensity of the reflected light. The photodiode converts the light signal into a weak current signal in the nA range. The AFE (Amplifier Element) amplifies the PPG signal through transimpedance, performs bandpass filtering from 0.5 to 40Hz, and suppresses baseline drift to remove motion noise, respiratory baseline drift, and power frequency interference. The signal is then converted into a digital signal by a 12-bit ADC with a sampling rate of 250Hz to ensure the integrity of the pulse wave's detailed characteristics.

[0040] Heart rate calculation employs an adaptive peak detection algorithm, identifying PPG peaks through a sliding window to eliminate false peaks caused by motion artifacts. The instantaneous heart rate, measured in beats per minute (BPM), is calculated by measuring the time interval between adjacent peaks, outputting once per second. HRV calculation uses a combined time-domain and frequency-domain approach: in the time domain, SDNN (standard deviation of all sinus intervals) and RMSSD (root mean square difference between adjacent intervals) are calculated; in the frequency domain, a 512-point Fast Fourier Transform is used to obtain low-frequency (LF, 0.04~0.15Hz) and high-frequency (HF, 0.15~0.4Hz) components, and the LF / HF ratio is calculated. A lower HRV and a higher LF / HF ratio indicate stronger sympathetic nerve excitation and higher emotional stress.

[0041] A 16-year-old female subject had a resting heart rate of 72 beats / minute, an RMSSD of 48 ms, and an LF / HF ratio of 1.2 in a calm and relaxed state. When she experienced self-harm impulses or emotional arousal, her heart rate rapidly increased to 108 beats / minute, her RMSSD decreased to 21 ms, and her LF / HF ratio increased to 3.1. The differences were significant and could effectively reflect changes in emotional stress.

[0042] Galvanic Skin Response (GSR), also known as electrical skin activity (EDA), reflects the secretory activity of sweat glands controlled by the sympathetic nervous system. When emotional stress increases, the sympathetic nervous system is excited, sweat gland secretion increases, skin conductivity increases, and resistance decreases. It is a classic indicator of emotional arousal and has the characteristics of fast response and high specificity. The response delay is usually less than 2 seconds, making it a core indicator for capturing sudden emotional arousal.

[0043] Using AC excitation to measure skin resistance avoids the electrode polarization problem caused by DC excitation. Specifically, DC current generates a polarization voltage at the electrode-skin interface, leading to measurement drift and electrode corrosion. AC excitation effectively avoids this problem and improves measurement stability. The AFE has a built-in programmable AC constant current source that outputs a sinusoidal AC current with a frequency of 10kHz and an amplitude of 10μA, which is injected into the skin through two excitation electrodes. Two measurement electrodes collect the voltage drop signal across the skin, which is input to an I / Q quadrature demodulation circuit to obtain the in-phase and quadrature components, respectively. The true value of skin resistance is calculated by the amplitude. The measurement range covers 1kΩ to 10MΩ, with a resolution of 10Ω and a sampling rate of 10Hz, ensuring the capture of rapid changes in skin resistance.

[0044] Simultaneously, skin resistance calculations incorporate subsequent skin temperature values ​​for temperature correction. A temperature compensation formula corrects for resistance drift caused by temperature changes, preventing misjudgments due to ambient temperature variations. Under normal, calm conditions, wrist skin resistance is approximately 100-500 kΩ. When emotional stress increases, sweat gland secretion increases, and resistance can drop by 20%-50% within seconds, a reaction faster than heart rate, making it a sensitive indicator of sudden emotional arousal.

[0045] The skin resistance of the subjects was 320kΩ when at rest; when aroused, it dropped rapidly to 140kΩ within 2 seconds, a decrease of 56%, which is much faster than the heart rate. This is the primary triggering characteristic for capturing sudden impulses.

[0046] Skin temperature reflects the state of peripheral blood circulation. When emotional stress increases, the excitation of the sympathetic nervous system leads to peripheral vasoconstriction and a decrease in skin temperature. At the same time, temperature changes affect the accuracy of skin resistance and PPG signals. Therefore, temperature measurement serves as an auxiliary indicator of emotional stress, a correction parameter for other physiological signals, and an important basis for the safe control of electrical stimulation.

[0047] A negative temperature coefficient thermistor (NTC) with a nominal resistance of 10kΩ and a B value of 3950K at 25℃ is used. It forms a voltage divider circuit with a 10kΩ precision low-temperature drift resistor. The AFE (Automatic Voltage Regulator) collects the voltage divider voltage, and the temperature is calculated using the Steinhart-Hart formula.

[0048] Where T is the Kelvin temperature. It is 298.15K (25℃). The resistance at the current temperature, The nominal resistance is set at 25℃, and B represents the thermistor value (material constant). The calculation accuracy reaches ±0.1℃, and the sampling rate is 1Hz, meeting the accuracy and speed requirements for temperature monitoring.

[0049] Skin temperature is used not only for mood assessment but also for the safe control of electrical stimulation: when skin temperature is below 15°C or above 40°C, the stimulation intensity is automatically reduced or stopped to avoid overstimulation due to decreased skin perception at low temperatures or the risk of skin damage at high temperatures. Simultaneously, a rapid increase in temperature may be caused by movement; combining this with the rate of temperature change can help identify movement states and suppress false triggers caused by movement.

[0050] The data is serialized in a fixed frame format, with each frame being 32 bytes long. It includes a frame header (2 bytes), a timestamp (4 bytes), heart rate (2 bytes), HRV parameters (RMSSD, LF / HF, a total of 4 bytes), skin resistance (4 bytes), skin temperature (2 bytes), and a checksum (2 bytes). The frame rate is 10Hz to ensure the real-time performance of the data.

[0051] The system employs an SPI serial peripheral interface for communication. The master device is a programmable electrical stimulation chip (STM32L432 low-power MCU), and the slave device is an AFE analog front-end. The communication rate is 2Mbps, and the short-distance board-level communication provides strong anti-interference capabilities. After the physiological feature dataset is transmitted to the programmable chip, it is stored in a 1KB data buffer for subsequent multi-parameter fusion calculations. Simultaneously, historical data from the most recent 30 seconds is cached for trend analysis to avoid misjudgments caused by single-point fluctuations. Each frame of data undergoes CRC verification; frames with incorrect verification results are discarded to ensure data reliability.

[0052] Furthermore, the step of performing multi-parameter fusion calculations on the physiological feature dataset using the programmable electrical stimulation chip to generate an emotional stress value includes: Four types of feature parameters—heart rate, heart rate variability, skin resistance, and skin temperature—are extracted from the physiological feature dataset, and the stress contribution value of each type of feature parameter is calculated based on a preset emotional stress assessment model. For feature parameters whose pressure contribution value is higher than the preset weight threshold, a weighted fusion calculation is performed to generate core pressure feature values; For feature parameters whose pressure contribution value is lower than the preset weight threshold, a correlation correction link with the core pressure feature value is established through a feature calibration algorithm. Based on the core stress characteristic value, and combined with the correction amount of the associated correction link, the corresponding emotional stress value is generated.

[0053] It should be noted that single physiological indicators are easily affected by factors such as exercise, temperature, and individual differences. Multi-parameter fusion can improve assessment accuracy through cross-validation. The emotional stress assessment model was trained based on clinical physiological data from 200 NSSI adolescents. It uses a standardized feature weighting method, with each feature parameter corresponding to a stress contribution value of 0 to 1, where a higher value indicates greater stress.

[0054] The stress contribution value is calculated using an individual baseline Z-score normalization method, rather than a universal fixed threshold. This is because the physiological baseline differences between individuals can exceed 30%, and a "one-size-fits-all" threshold would lead to false triggering in individuals with high baselines and missed triggering in individuals with low baselines. When a user first wears the device, a 3-day baseline calibration is performed, collecting physiological data for 5 minutes each day in a seated, calm state, and calculating the individual baseline mean and standard deviation. During real-time calculation, the current feature value is compared with the baseline, converted into a Z-score, and then mapped to a stress contribution value of 0-1 using a Sigmoid function to accommodate individual differences.

[0055] The basic weights for the four features are: heart rate 0.25, HRV (RMSSD+LF / HF) 0.35, skin resistance 0.25, and skin temperature 0.15. HRV has the highest weight because it has the highest specificity for emotional stress and is relatively less affected by factors such as exercise; skin resistance reacts the fastest and is suitable for capturing sudden emotional arousal; skin temperature serves as an auxiliary correction indicator. The weights can be dynamically adjusted according to the scenario. For example, in exercise detection mode, the heart rate weight can be reduced to 0.1, while the HRV weight can be increased to 0.5 to avoid misjudgments caused by increased heart rate due to exercise.

[0056] A subject's current heart rate is 95 beats / min, baseline 72 beats / min, Z-score 2.1, and mapped stress contribution value 0.82; RMSSD is 25ms, baseline 48ms, Z-score -2.3, and stress contribution value 0.88; skin resistance is 160kΩ, baseline 320kΩ, Z-score -1.8, and stress contribution value 0.76; skin temperature is 32.5℃, baseline 33.2℃, Z-score -1.1, and stress contribution value 0.65.

[0057] The preset weight threshold is 0.5. When the pressure contribution value of a feature is higher than 0.5, it is judged as a "pressure-related feature" and included in the weighted fusion of core features; features with a value lower than 0.5 are "weakly related features" and do not directly participate in the core calculation, but are corrected through the correlation correction link. This mechanism can avoid noise interference from low contribution value features and improve the stability of the evaluation.

[0058] The core pressure characteristic values ​​are calculated using a weighted summation:

[0059] in Normalized weights for core features This represents the corresponding stress contribution value. The core stress characteristic value reflects the level of emotional stress as reflected by the main physiological indicators and is the core basis for assessment.

[0060] Taking the above example, the contribution values ​​of heart rate, HRV, and skin resistance are all higher than 0.5, and are included in the core calculation, with normalized weights of 0.3, 0.4, and 0.3, respectively; the contribution value of skin temperature is 0.65, which is also higher than the threshold, with a weight of 0.15. The core stress value = 0.3×0.82 + 0.4×0.88 + 0.3×0.76 + 0.15×0.65 = 0.246 + 0.352 + 0.228 + 0.0975 = 0.9235, reflecting the level of severe emotional stress.

[0061] While secondary features contribute little to stress levels, they can be used to validate the effectiveness of core features and eliminate confounding factors. For example, a high core stress value coupled with a decrease in skin temperature can confirm that the vasoconstriction is caused by emotional stress; however, if a high core stress value is accompanied by an increase in skin temperature and a rise in heart rate with motion detected by an accelerometer, it may be caused by exercise, and the core stress value needs to be corrected.

[0062] When the trends of secondary features and core features are consistent, the correction coefficient is 1, and the pressure value remains unchanged; when the trends are inconsistent, the correction coefficient is less than 1, reducing the core pressure value and avoiding misjudgment. The correction coefficient ranges from 0.5 to 1.0 and is dynamically adjusted according to the degree of interference.

[0063] During the subject's walking, heart rate increased to 90 beats / minute (contribution value 0.7), HRV was 40 ms (contribution value 0.4), skin resistance was 300 kΩ (contribution value 0.3), and skin temperature increased by 0.5℃, while the accelerometer detected continuous movement. At this point, the initial calculated core pressure value was 0.55. Considering the movement characteristics and the trend of temperature increase, a correction factor of 0.6 was applied, resulting in a corrected pressure value of 0.33, which is below the trigger threshold of 0.6, thus preventing false triggering during movement.

[0064] Based on core stress characteristics, and combined with the correction amount of the associated correction link, an emotional stress value is generated. The emotional stress value ranges from 0 to 1, divided into four levels: 0-0.3 is normal, 0.3-0.6 is mild stress, 0.6-0.8 is moderate stress, and 0.8-1.0 is severe stress. Different levels correspond to different intervention strategies, achieving tiered intervention.

[0065] Simultaneously, a duration condition is set: only when the stress value exceeds the threshold and lasts for more than 3 seconds is it considered a valid trigger, avoiding false stimulation caused by instantaneous fluctuations. For example, signal fluctuations caused by a single blink or arm movement, lasting less than 2 seconds, will not trigger intervention; only when the stress value remains above the threshold is it determined that emotional stress is continuously increasing and there is a risk of self-harm impulses, thus initiating intervention.

[0066] The trigger detection also incorporates scene modes: in normal mode, moderate stress lasting 5 seconds triggers the trigger; in exercise mode, severe stress lasting 10 seconds triggers the trigger; and in sleep mode, a dual threshold of skin resistance and HRV is used to avoid false triggering due to physiological fluctuations during sleep. These multiple modes adapt to different life scenarios, further reducing the false trigger rate.

[0067] Furthermore, the step of generating an electrical stimulation control signal by the programmable electrical stimulation chip and outputting an electrical stimulation current of preset parameters to the human body through dry electrodes in contact with the human body when the intervention trigger signal is in an effective state includes: The intervention trigger signals are labeled with their types to distinguish between single emergency triggers, continuous high-pressure triggers, and abnormal triggers, and the duration and frequency of each type of trigger are statistically analyzed. Based on the type and duration of the trigger, the electrical stimulation output mode is determined, and the key parameters of frequency, pulse width, and current intensity contained in the electrical stimulation output mode are extracted. Several output control sub-items are then divided according to the priority of the key parameters. Based on the hierarchical relationship between the electrical stimulation output mode and several output control sub-items, an electrical stimulation control sequence is constructed, and the output channel ID of the dry electrode is marked at each node of the electrical stimulation control sequence to generate the corresponding electrical stimulation control signal.

[0068] It should be noted that different trigger types correspond to different intervention modes, accurately matching different characteristics of self-harm impulses to achieve "intervention on demand" and avoid overstimulation or understimulation.

[0069] Criteria and scenarios for determining three types of triggers: A single emergency trigger: The emotional stress value rises from below 0.5 to above 0.8 within 3 seconds and lasts for 5 to 30 seconds. This corresponds to a sudden self-harm impulse, which is usually triggered by a specific event or thought. The impulse is intense and short-lived, requiring rapid and intense intervention to break the impulse chain.

[0070] Sustained high-pressure trigger: Emotional stress value remains above 0.6 for more than 2 minutes and fluctuates less than 0.2, corresponding to persistent anxiety, irritability, and recurring self-harm impulses. Continuous and gradual intervention is needed to relieve the sustained high-stress state, while avoiding skin adaptation caused by continuous strong stimulation.

[0071] Abnormal Trigger: If the pressure value rises abnormally to above 0.9, accompanied by abnormal physiological signals such as heart rate exceeding 180 beats / minute, skin resistance below 500Ω or above 10MΩ, or skin temperature below 10℃ or above 42℃, it may be caused by equipment failure, electrode detachment, or strenuous exercise. No stimulation will be output, and the device will enter safety mode. The screen will prompt the user to check the device.

[0072] Simultaneously, the duration of triggers, trigger frequency, and daily cumulative stimulation time are statistically analyzed to evaluate the intervention effect and user habits, and also serve as a basis for safety control: for example, if more than 5 triggers occur within 1 hour, the maximum stimulation intensity is automatically reduced by 20% to avoid overstimulation; if the cumulative stimulation time in 24 hours exceeds 30 minutes, the automatic intervention function is automatically paused to prompt the user to rest and avoid discomfort caused by prolonged electrical stimulation.

[0073] The electrical stimulation uses a constant current pulse output, falling under the category of transcutaneous electrical nerve stimulation (TENS). It stimulates the sensory fibers of the median and ulnar nerves via wrist electrodes, modulating nerve excitability and simultaneously promoting the release of endorphins to alleviate impulses and pain. All stimulation parameters comply with the IEC60601-1 safety standard for medical electrical equipment, with a maximum current not exceeding 2mA, falling within the microcurrent range, posing no risk of electric shock or permanent damage.

[0074] It should be noted that the intervention system provided in this embodiment needs to be set up and used by professional medical personnel under formal medical supervision. Furthermore, the maximum current set in this embodiment is locked by the system and cannot be modified by the user. The maximum current will never exceed 2mA.

[0075] Additionally, it should be noted that for therapeutic electrical stimulation devices, IEC 60601-2-10 clearly specifies the upper limit of output current under a 500Ω load: DC ≤80mA, AC ≤50mA for frequencies ≤400Hz, with higher limits for higher frequencies. A maximum stimulation current of 2mA is only 2.5% (DC scenario) to 4% (low-frequency AC scenario) of the standard limit, far below the safety threshold, and fully meets the mandatory requirements for output parameters.

[0076] Different trigger types correspond to different output modes, and the parameters have been optimized through pre-clinical trials: Single emergency trigger: Employs a dual-stage mode of "strong stimulation + slow release". The first stage is the strong stimulation period, which outputs a symmetrical biphasic pulse with a current of 1.5mA, a frequency of 10Hz, and a pulse width of 200μs for 5 seconds to quickly activate neural regulation and interrupt impulsive thinking. The second stage is the slow release period, where the current drops to 0.8mA and the frequency drops to 5Hz for 10 seconds to slow down emotions and avoid impulsive rebound caused by sudden cessation.

[0077] Continuous high-pressure triggering: The "progressive stimulation" mode is adopted, starting from 0.5mA and increasing by 0.2mA every 10 seconds, with a maximum of 2mA. At the same time, the frequency gradually increases from 8Hz to 15Hz, and is dynamically adjusted according to the real-time pressure value: the intensity is reduced when the pressure decreases and gradually increased when the pressure continues to rise, so as to ensure the effect and avoid the neural adaptation caused by continuous strong stimulation.

[0078] Abnormal trigger: When the output current is 0mA, stimulation will stop immediately, and the screen will display "Signal abnormal, please check the wearing position". At the same time, an abnormal log will be recorded for subsequent analysis.

[0079] The output control sub-items are divided into three priority levels: the first priority is current intensity, which directly relates to safety and performance; the second priority is frequency; and the third priority is pulse width and duration. This priority division ensures that core parameters are controlled first, and that current intensity remains within a safe range during parameter adjustments to avoid safety issues.

[0080] The timing control defines the timing nodes of the stimulus, the parameter change nodes, and the channel switching nodes, ensuring the timing accuracy and channel correctness of the stimulus output.

[0081] Taking a single emergency trigger as an example, the control sequence is divided into 4 stages, each with specific node parameters: Pre-stimulation detection phase (0~1s): Output 0.3mA test current to detect skin impedance. If the impedance is within the normal range of 500Ω~50kΩ, proceed to main stimulation. If the impedance is below 500Ω (possibly a short circuit) or above 50kΩ (possibly electrode detachment), stop stimulation and provide a warning to ensure safety.

[0082] Strong stimulation phase (1~6s): Output 1.5mA, 10Hz, 200μs pulse, channel 1 output, skin impedance is detected once per second.

[0083] Sustained-release phase (6~16s): The current is reduced to 0.8mA and the frequency is reduced to 5Hz. Channel 1 and Channel 2 alternate output, switching every 2 seconds to avoid local skin nerve adaptation and enhance the sustained effect.

[0084] Stop phase (16~17s): The current drops linearly to 0 for 1 second to avoid the stinging sensation caused by sudden current changes and improve comfort.

[0085] Each node controlling the timing is marked with a channel ID, current value, and duration. The programmable chip outputs node by node according to the timing sequence, ensuring the accuracy of the stimulation. At the same time, skin impedance is detected at each node, and stimulation is immediately stopped if the impedance is abnormal to ensure safety.

[0086] Furthermore, the step of marking the output channel ID of the dry electrode at each node of the electrical stimulation control sequence to generate the corresponding electrical stimulation control signal includes: Each node of the electrical stimulation control timing sequence is hierarchically encoded to form a three-dimensional control code that includes stimulation stage identifiers, timing node identifiers, and output channel identifiers. The output channel IDs of the dry electrode are grouped and clustered according to their output polarity, and a mapping relationship between each output channel ID and the three-dimensional control code is constructed. According to the mapping relationship, the output channel ID is written into the control field of each node of the electrical stimulation control sequence, and the current intensity value and duration weight value of the node are marked next to each output channel ID to generate the corresponding electrical stimulation control signal.

[0087] It should be noted that the three-dimensional control code encodes information from three dimensions—stimulation stage, timing position, and output channel—into a single control word. The programmable chip can obtain all control parameters for that node by parsing the control code, thereby improving control efficiency and reducing the probability of errors.

[0088] The control code uses a 16-bit fixed-length encoding format, with the specific bit allocation as follows: The first three digits indicate the stimulation phase, with a total of 8 phase codes, such as 001=prestimulation, 010=main stimulation, 011=slow release, 100=stop, 101=abnormal stop, etc. The 4th to 9th positions are time sequence node identifiers, with a total of 64 nodes, corresponding to 0 to 63 time steps. The time step size is 100ms, and the longest time sequence can be 6.4 seconds. Longer time sequences are achieved by splicing multiple frames. Bits 10-13: Output channel identifier, a total of 16 channel codes, supporting dual-channel output of up to 8 electrodes, expandable; Bits 14-16: Odd parity bits, used for transmission and storage verification to avoid abnormal output caused by control bit errors.

[0089] For example, at the 5th time step (500ms) of the main stimulus phase, the control code output by channel 1 is: 01000010100011. After parsing, the phase, time, and channel information can be obtained. The three-dimensional encoding ensures the structure of the control signal, which is convenient for storage and parsing. At the same time, the reliability is ensured by the check bit, and the error rate of the control code is less than 0.01%.

[0090] The wristband is equipped with four medical-grade dry electrodes made of silver-plated silver chloride (Ag / AgCl), which have good biocompatibility and low contact resistance. The electrodes are divided into two groups of output channels according to their positions. Each group contains a positive electrode and a negative electrode, which can realize single-channel output, dual-channel alternating output, and dual-channel synchronous output.

[0091] The positions of the four electrodes are designed according to the nerve distribution in the wrist: channel 1 corresponds to the electrode on the upper medial side of the wrist (median nerve course), channel 2 corresponds to the electrode on the lower medial side of the wrist (ulnar nerve course), and channels 3 and 4 are lateral reference electrodes. Stimulation at different positions can activate different nerve fibers, and the alternating output of the dual channels can avoid local nerve adaptation and prolong the effective stimulation time.

[0092] The mapping relationship is stored in the chip's lookup table (LUT). Each channel ID corresponds to a set of electrode switch control bits, including positive switch, negative switch, current source selection, and polarity reversal control. After the control code is parsed, the corresponding channel control bit is quickly found through the mapping table to control the analog switch switching and achieve channel selection. The mapping relationship also supports polarity reversal function, automatically switching between positive and negative every 5 seconds to avoid electrode polarization, extend electrode life, and reduce skin irritation.

[0093] The control signal consists of three parts: timing control, channel control, and current control. The PWM peripheral of the programmable chip outputs pulses with corresponding frequency, pulse width, and duty cycle according to the control signal, and outputs them to the corresponding electrodes through the Howland constant current output circuit.

[0094] The current intensity value uses 12-bit encoding, corresponding to an output range of 0~2mA, with a resolution of approximately 0.5μA, ensuring fine-grained current adjustment. The duration weight value is used to calculate the duration of the node, combined with the timing node identifier, to achieve precise time control. After the control signal is generated, it undergoes triple safety checks: whether the current exceeds the maximum value, whether the channel is valid, and whether the timing is continuous. Only after passing these checks is the signal output to avoid erroneous stimulation.

[0095] The constant current output circuit uses an improved Howland current source with an output impedance greater than 10MΩ. This ensures that when the skin impedance varies within the range of 500Ω to 50kΩ, the output current fluctuation is less than 1%, achieving true constant current stimulation and avoiding sudden changes in current caused by changes in skin impedance, thus ensuring consistent stimulation effect and safety.

[0096] Furthermore, the steps of displaying the wristband's working status data on the screen and collecting the user's stop command via the button switch include: The physiological feature dataset, emotional stress value, and electrical stimulation state data are classified to generate a state dataset containing working mode, physiological indicators, stimulation state, and device power. The input signal of the button switch is analyzed by edge detection to generate a trigger instruction set. The state dataset is hierarchically matched with the display partitions of the display screen to determine the display priority of each display partition. The trigger instruction set is compared with the preset control instructions by attributes, and valid instructions with a matching degree exceeding the preset matching threshold are retained. High-priority status data is formatted for display to generate display frame data, and valid instructions are parsed to generate a set of control instructions. The display frame data and the set of control instructions are then integrated to generate interactive control signals.

[0097] It should be noted that the state dataset is divided into four main categories and is updated in real time: Working modes: Normal mode, sports mode, sleep mode, and manual mode. The pressure threshold and stimulation parameters are different in different modes, and users can switch between them by pressing a button.

[0098] Physiological indicators: heart rate, stress level, skin resistance, and skin temperature, updated every second and displayed intuitively using numbers and progress bars.

[0099] Stimulation status: standby, stimulating, paused, abnormal, displaying the current stimulation intensity, remaining time, and channel status.

[0100] Device status: battery level, Bluetooth connection, wearing status, and error messages.

[0101] The push-button switch uses two physical tactile buttons: a function button and a stop button. It has an IP68 waterproof rating and supports operation while worn. The button signals are stabilized by both hardware and software, with a 20ms stabilization time. Only the rising edge of the pressed button is recognized to prevent accidental triggering. The stop button has the highest priority; pressing it in any state immediately stops electrical stimulation, ensuring the user can terminate the intervention at any time and respecting user autonomy. The function button supports short presses to switch modes, long presses for 3 seconds to manually trigger stimulation, and double-clicks to view historical data, meeting the user's need for active intervention.

[0102] After key parsing, a set of trigger commands is generated, including stop commands, mode switching commands, manual trigger commands, history query commands, etc. Each command has a priority label, with the stop command having the highest priority (level 0), which can interrupt any ongoing automatic stimulation.

[0103] The bracelet features a 0.96-inch OLED display with a resolution of 160×80, divided into four display zones arranged from highest to lowest priority: Top status bar: Displays current mode, battery level, and connection status; always displayed. Central main display area: Displays pressure level (progress bar) and heart rate value; during stimulation, it switches to stimulation intensity and remaining time. Lower status area: Displays skin resistance, temperature, and stimulation channel status; Bottom prompt area: Displays error messages and operation tips.

[0104] The display priority mechanism ensures that important information is displayed first: when there is a stimulus, the main display area prioritizes the display of stimulus intensity and remaining time; when there is an abnormality, the abnormality prompt is displayed in full screen first, ensuring that users see key information as soon as possible.

[0105] The triggered command is compared with the preset command library. If the command signature and checksum match, it is considered a valid command; otherwise, it is discarded. This avoids erroneous commands caused by static interference or key bounce. For example, the signature of a stop command is 0x55. If the checksum matches, the stop operation is executed; otherwise, it is ignored, improving the reliability of command recognition.

[0106] The display frame data is formatted according to the OLED display format, including pixel data, refresh address, and refresh rate, refreshing 10 times per second to ensure real-time status display. The control instruction set contains all parsed valid instructions, sorted by priority, with higher priority instructions executed first. The level 0 stop instruction can interrupt any level 1 or 2 operation.

[0107] The integration of display and control ensures the synchronization of status display and control actions, avoiding inconsistencies between the displayed status and actual control. For example, when a user presses the stop button, the instruction is interpreted as stopping the stimulus, which has the highest priority and is executed immediately; at the same time, the display status is updated from "stimulating" to "stopped". The display and control are completed synchronously, and the user can immediately see the status change, providing clear interactive feedback.

[0108] Furthermore, the step of integrating the display frame data and the control command set to generate an interactive control signal includes: Add display timing tags to the display frame data and add appropriate instruction priority tags to the control instruction set; An interactive control index table is created based on the display timing tag and the instruction priority tag, and a state association matrix is ​​generated based on the display frame data and the control instruction set. The control parameters are extracted from the state association matrix according to the interactive control index table, and the control parameters are set as interactive control signals.

[0109] It should be noted that the display time stamp is a 16-bit timestamp with a precision of 10ms. Each frame of displayed data has a unique timestamp, and the display screen refreshes in the order of timestamps, with new data overwriting old data to avoid old data overwriting new data.

[0110] Instruction priority is divided into three levels: Level 0 (highest): Stop command, abnormal stop command, can interrupt any operation and execute immediately; Level 1 (Medium): Mode switching command, manual trigger command, executed after the current operation is completed; Level 2 (lowest): Automatic stimulation command and status update command are executed in sequence.

[0111] The priority mechanism ensures the highest priority of safety commands. Under any circumstances, pressing the stop button will immediately terminate the stimulation, while ensuring that status updates do not affect high-priority controls.

[0112] The interactive control index table is an index structure for display and control, storing the display data address and control command address at each point in time. It allows for fast addressing based on time, improving processing efficiency.

[0113] The state association matrix is ​​a two-dimensional matrix, with rows corresponding to time nodes and columns corresponding to state types. Each cell stores the display value and control value at that time point, with a one-to-one correspondence between the two. After each control action is executed, the state association matrix is ​​updated synchronously, and then the display is refreshed to ensure consistency between "control-state-display". For example, after the "stop stimulus" command is executed, the stimulus state in the matrix is ​​updated to "stop", and the display frame is updated simultaneously. The state seen by the user is completely synchronized with the actual control state, and there will be no inconsistency problem of "the stimulus has stopped but the display is still showing stimulation".

[0114] The status correlation matrix also stores historical data, which can record stress changes and stimulation records over the past 7 days. It can be synchronized to a mobile APP via Bluetooth for guardians to view and doctors to evaluate.

[0115] Five safety mechanisms are embedded in the interactive control signals to ensure safe intervention: Timeout stop mechanism: Each automatic stimulation session lasts a maximum of 30 seconds, after which it will automatically stop to avoid prolonged stimulation; the cumulative stimulation time per day shall not exceed 30 minutes, otherwise the automatic function will be suspended.

[0116] Abnormal impedance: Stop stimulation immediately if skin impedance is below 500Ω (short circuit) or above 50kΩ (electrode detachment), and check the wearer.

[0117] Key priority mechanism: At any time, the stop key command takes precedence over all automatic controls, and the user can terminate the system at any time.

[0118] Dual current limiting: The software sets the maximum current to 2mA, while the hardware constant current circuit outputs a maximum of 2.2mA, ensuring safety through dual current limiting.

[0119] Temperature protection: Reduce or stop stimulation when skin temperature is below 15℃ or above 40℃ to avoid skin damage caused by abnormal temperature.

[0120] Please see Figure 3 The third embodiment of the present invention provides: A non-suicidal self-harm intervention system based on an electrical stimulation wristband, wherein the system comprises: The processing module is used to collect physiological signals from the wrist via the built-in AFE of the wristband, and to condition, demodulate and convert the excitation current and induced current signals during the collection process to generate a physiological feature dataset. The physiological feature dataset is then transmitted to the programmable electrical stimulation chip via SPI communication. The calculation module is used to perform multi-parameter fusion calculation on the physiological feature dataset through the programmable electrical stimulation chip to generate an emotional stress value, and compare the emotional stress value with a preset emotional stress threshold to obtain an intervention trigger signal; The intervention module is used to generate an electrical stimulation control signal when the intervention trigger signal is in an effective state, and output an electrical stimulation current with preset parameters to the human body through dry electrodes in contact with the human body in order to perform active intervention for non-suicidal self-harm behavior. The control module is used to display the wristband's working status data on the screen and to collect user stop commands via button switches, thereby controlling the start and stop of the electrical stimulation current output.

[0121] Furthermore, the steps of acquiring physiological signals from the wrist via the built-in AFE of the wristband, and conditioning, demodulating, and performing analog-to-digital conversion on the excitation current and induced current signals during the acquisition process include: Heart rate light-sensing signals are collected and converted into current signals, which are then conditioned by AFE and converted into digital signals to calculate heart rate and heart rate variability indices. A weak AC excitation current with preset parameters is injected into the skin by an AC constant current source, the voltage drop signal across the skin is measured, and the voltage drop signal is demodulated and calculated by the I / Q demodulation circuit inside the AFE to obtain the skin resistance value. The resistance signal of the NTC is collected and calculated using a voltage divider circuit and the B-value formula to obtain the skin temperature value. Heart rate, heart rate variability, skin resistance, and skin temperature data are serialized to generate a physiological feature dataset, which is then transmitted to a programmable electrical stimulation chip via SPI communication.

[0122] Furthermore, the step of performing multi-parameter fusion calculations on the physiological feature dataset using the programmable electrical stimulation chip to generate an emotional stress value includes: Four types of feature parameters—heart rate, heart rate variability, skin resistance, and skin temperature—are extracted from the physiological feature dataset, and the stress contribution value of each type of feature parameter is calculated based on a preset emotional stress assessment model. For feature parameters whose pressure contribution value is higher than the preset weight threshold, a weighted fusion calculation is performed to generate core pressure feature values; For feature parameters whose pressure contribution value is lower than the preset weight threshold, a correlation correction link with the core pressure feature value is established through a feature calibration algorithm. Based on the core stress characteristic value, and combined with the correction amount of the associated correction link, the corresponding emotional stress value is generated.

[0123] Furthermore, the step of generating an electrical stimulation control signal by the programmable electrical stimulation chip and outputting an electrical stimulation current of preset parameters to the human body through dry electrodes in contact with the human body when the intervention trigger signal is in an effective state includes: The intervention trigger signals are labeled with their types to distinguish between single emergency triggers, continuous high-pressure triggers, and abnormal triggers, and the duration and frequency of each type of trigger are statistically analyzed. Based on the type and duration of the trigger, the electrical stimulation output mode is determined, and the key parameters of frequency, pulse width, and current intensity contained in the electrical stimulation output mode are extracted. Several output control sub-items are then divided according to the priority of the key parameters. Based on the hierarchical relationship between the electrical stimulation output mode and several output control sub-items, an electrical stimulation control sequence is constructed, and the output channel ID of the dry electrode is marked at each node of the electrical stimulation control sequence to generate the corresponding electrical stimulation control signal.

[0124] Furthermore, the step of marking the output channel ID of the dry electrode at each node of the electrical stimulation control sequence to generate the corresponding electrical stimulation control signal includes: Each node of the electrical stimulation control timing sequence is hierarchically encoded to form a three-dimensional control code that includes stimulation stage identifiers, timing node identifiers, and output channel identifiers. The output channel IDs of the dry electrode are grouped and clustered according to their output polarity, and a mapping relationship between each output channel ID and the three-dimensional control code is constructed. According to the mapping relationship, the output channel ID is written into the control field of each node of the electrical stimulation control sequence, and the current intensity value and duration weight value of the node are marked next to each output channel ID to generate the corresponding electrical stimulation control signal.

[0125] Furthermore, the steps of displaying the wristband's working status data on the screen and collecting the user's stop command via the button switch include: The physiological feature dataset, emotional stress value, and electrical stimulation state data are classified to generate a state dataset containing working mode, physiological indicators, stimulation state, and device power. The input signal of the button switch is analyzed by edge detection to generate a trigger instruction set. The state dataset is hierarchically matched with the display partitions of the display screen to determine the display priority of each display partition. The trigger instruction set is compared with the preset control instructions by attributes, and valid instructions with a matching degree exceeding the preset matching threshold are retained. High-priority status data is formatted for display to generate display frame data, and valid instructions are parsed to generate a set of control instructions. The display frame data and the set of control instructions are then integrated to generate interactive control signals.

[0126] Furthermore, the step of integrating the display frame data and the control command set to generate an interactive control signal includes: Add display timing tags to the display frame data and add appropriate instruction priority tags to the control instruction set; An interactive control index table is created based on the display timing tag and the instruction priority tag, and a state association matrix is ​​generated based on the display frame data and the control instruction set. The control parameters are extracted from the state association matrix according to the interactive control index table, and the control parameters are set as interactive control signals.

[0127] The fourth embodiment of the present invention provides an electrical stimulation bracelet, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the functions of the non-suicidal self-harm behavior intervention system based on the electrical stimulation bracelet as described above.

[0128] The fifth embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the functions of the non-suicidal self-harm behavior intervention system based on an electrical stimulation bracelet as described above.

[0129] In summary, the above embodiments of the present invention can generate an electrical stimulation control signal in real time when a self-harm impulse occurs, and output an electrical stimulation current with preset parameters to the human body through dry electrodes, thereby realizing active and real-time intervention in non-suicidal self-harm behavior; at the same time, the working status is displayed on the screen and the stop command is collected by the button switch, supporting the user to actively control the start and stop of the intervention, effectively improving the timeliness, accuracy and safety of the intervention.

[0130] It should be noted that the above modules can be functional modules or program modules, and can be implemented in software or hardware. Furthermore, these modules communicate with each other via communication interfaces and buses to transmit and process signals. For modules implemented in hardware, these modules can reside in the same processor; alternatively, they can be located in different processors in any combination.

[0131] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0132] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0133] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0134] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0135] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A non-suicidal self-harm intervention system based on an electrical stimulation bracelet, characterized in that, The system includes: The processing module is used to collect physiological signals from the wrist via the built-in AFE of the wristband, and to condition, demodulate and convert the excitation current and induced current signals during the collection process to generate a physiological feature dataset. The physiological feature dataset is then transmitted to the programmable electrical stimulation chip via SPI communication. The calculation module is used to perform multi-parameter fusion calculation on the physiological feature dataset through the programmable electrical stimulation chip to generate an emotional stress value, and compare the emotional stress value with a preset emotional stress threshold to obtain an intervention trigger signal; The intervention module is used to generate an electrical stimulation control signal when the intervention trigger signal is in an effective state, and output an electrical stimulation current with preset parameters to the human body through dry electrodes in contact with the human body in order to perform active intervention for non-suicidal self-harm behavior. The control module is used to display the wristband's working status data on the screen and to collect the user's stop command through the button switch, so as to control the start and stop of the electrical stimulation current output accordingly.

2. The non-suicidal self-harm intervention system based on an electrical stimulation bracelet according to claim 1, characterized in that, The steps of acquiring physiological signals from the wrist via the built-in AFE of the wristband, and conditioning, demodulating, and performing analog-to-digital conversion on the excitation current and induced current signals during the acquisition process include: Heart rate light-sensing signals are collected and converted into current signals, which are then conditioned by AFE and converted into digital signals to calculate heart rate and heart rate variability indices. A weak AC excitation current with preset parameters is injected into the skin by an AC constant current source, the voltage drop signal across the skin is measured, and the voltage drop signal is demodulated and calculated by the I / Q demodulation circuit inside the AFE to obtain the skin resistance value. The resistance signal of the NTC is collected and calculated using a voltage divider circuit and the B-value formula to obtain the skin temperature value. Heart rate, heart rate variability, skin resistance, and skin temperature data are serialized to generate a physiological feature dataset, which is then transmitted to a programmable electrical stimulation chip via SPI communication.

3. The non-suicidal self-harm intervention system based on an electrical stimulation bracelet according to claim 2, characterized in that, The step of performing multi-parameter fusion calculations on the physiological feature dataset using the programmable electrical stimulation chip to generate an emotional stress value includes: Four types of feature parameters—heart rate, heart rate variability, skin resistance, and skin temperature—are extracted from the physiological feature dataset, and the stress contribution value of each type of feature parameter is calculated based on a preset emotional stress assessment model. For feature parameters whose pressure contribution value is higher than the preset weight threshold, a weighted fusion calculation is performed to generate core pressure feature values; For feature parameters whose pressure contribution value is lower than the preset weight threshold, a correlation correction link with the core pressure feature value is established through a feature calibration algorithm. Based on the core stress characteristic value, and combined with the correction amount of the associated correction link, the corresponding emotional stress value is generated.

4. The non-suicidal self-harm intervention system based on an electrical stimulation bracelet according to claim 1, characterized in that, The step of generating an electrical stimulation control signal by the programmable electrical stimulation chip and outputting an electrical stimulation current with preset parameters to the human body through dry electrodes in contact with the human body when the intervention trigger signal is valid includes: The intervention trigger signals are labeled with their types to distinguish between single emergency triggers, continuous high-pressure triggers, and abnormal triggers, and the duration and frequency of each type of trigger are statistically analyzed. Based on the type and duration of the trigger, the electrical stimulation output mode is determined, and the key parameters of frequency, pulse width, and current intensity contained in the electrical stimulation output mode are extracted. Several output control sub-items are then divided according to the priority of the key parameters. Based on the hierarchical relationship between the electrical stimulation output mode and several output control sub-items, an electrical stimulation control sequence is constructed, and the output channel ID of the dry electrode is marked at each node of the electrical stimulation control sequence to generate the corresponding electrical stimulation control signal.

5. The non-suicidal self-harm intervention system based on an electrical stimulation bracelet according to claim 4, characterized in that, The step of marking the output channel ID of the dry electrode at each node of the electrical stimulation control sequence to generate the corresponding electrical stimulation control signal includes: Each node of the electrical stimulation control timing sequence is hierarchically encoded to form a three-dimensional control code that includes stimulation stage identifiers, timing node identifiers, and output channel identifiers. The output channel IDs of the dry electrode are grouped and clustered according to their output polarity, and a mapping relationship between each output channel ID and the three-dimensional control code is constructed. According to the mapping relationship, the output channel ID is written into the control field of each node of the electrical stimulation control sequence, and the current intensity value and duration weight value of the node are marked next to each output channel ID to generate the corresponding electrical stimulation control signal.

6. The non-suicidal self-harm intervention system based on an electrical stimulation bracelet according to claim 1, characterized in that, The steps of displaying the wristband's working status data on the screen and collecting the user's stop command via the button switch include: The physiological feature dataset, emotional stress value, and electrical stimulation state data are classified to generate a state dataset containing working mode, physiological indicators, stimulation state, and device power. The input signal of the button switch is analyzed by edge detection to generate a trigger instruction set. The state dataset is hierarchically matched with the display partitions of the display screen to determine the display priority of each display partition. The trigger instruction set is compared with the preset control instructions by attributes, and valid instructions with a matching degree exceeding the preset matching threshold are retained. High-priority status data is formatted for display to generate display frame data, and valid instructions are parsed to generate a set of control instructions. The display frame data and the set of control instructions are then integrated to generate interactive control signals.

7. The non-suicidal self-harm intervention system based on an electrical stimulation bracelet according to claim 6, characterized in that, The step of integrating the display frame data and the control command set to generate interactive control signals includes: Add display timing tags to the display frame data and add appropriate instruction priority tags to the control instruction set; An interactive control index table is created based on the display timing tag and the instruction priority tag, and a state association matrix is ​​generated based on the display frame data and the control instruction set; The control parameters are extracted from the state association matrix according to the interactive control index table, and the control parameters are set as interactive control signals.

8. An electrical stimulation wristband, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the functions of the non-suicidal self-harm intervention system based on an electrical stimulation bracelet as described in any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the functionality of the non-suicidal self-harm intervention system based on an electrical stimulation bracelet as described in any one of claims 1 to 7.

Images