Closed-loop nerve regulation and control system for hypertension intervention

Abstract

The invention discloses a closed-loop nerve regulation and control system for hypertension intervention, and belongs to the technical field of medical electronic equipment. The system comprises a blood pressure monitoring module, a transcranial alternating current stimulation module, a trigeminal nerve activation monitoring module and a control unit, and is configured to continuously obtain blood pressure data of a user through the blood pressure monitoring module and run an intervention program when the blood pressure data is judged to be hypertension data; the intervention procedure comprises: instructing the transcranial alternating current stimulation module to apply therapeutic transcranial alternating current stimulation to the user; during the period of applying the therapeutic transcranial alternating current stimulation, instructing the trigeminal nerve activation monitoring module to obtain a nerve activation signal, and taking the nerve activation signal as nerve feedback; instructing the blood pressure monitoring module to continuously obtain the blood pressure data of the user, and taking the blood pressure data as physiological feedback; and adjusting stimulation parameters of the transcranial alternating current stimulation module according to the neural feedback and the physiological feedback. And closed-loop nerve regulation for hypertension intervention is realized.

Description

Technical Field

[0001] This invention belongs to the field of medical electronic equipment technology, and in particular relates to a closed-loop neural modulation system for hypertension intervention. Background Technology

[0002] Hypertension is the leading risk factor for cardiovascular disease and premature death worldwide, and its effective management is crucial for public health. Traditional blood pressure management relies on medication and regular cuffed blood pressure measurements, which fail to capture the diurnal rhythm and transient fluctuations of blood pressure, thus limiting the comprehensive assessment and management of hypertension. More advanced monitoring and intervention technologies are needed.

[0003] Current technologies have significant gaps in continuous blood pressure monitoring, neuromodulation intervention, and the combination of the two. On the one hand, cuffless blood pressure monitoring suffers from model drift due to physiological factors such as heart rate and autonomic nervous system status, resulting in low measurement accuracy and poor reliability, and requiring frequent manual calibration. On the other hand, neuromodulation intervention typically operates in an open-loop mode, lacking real-time feedback on the neural activity status of the treatment target, leading to a lack of personalized treatment plans and low efficacy and efficiency. Summary of the Invention

[0004] To address the aforementioned problems in the existing technology, the purpose of this invention is to provide a closed-loop neuromodulation system for hypertension intervention. This system solves the problems of poor efficacy of open-loop stimulation and inaccurate blood pressure monitoring without a cuff in the existing technology. The technical solution adopted is as follows:

[0005] A closed-loop neuromodulation system for hypertension intervention includes: a blood pressure monitoring module configured to continuously monitor a user's blood pressure and output blood pressure data representing blood pressure levels for acquiring the user's blood pressure data; a trigeminal nerve activation monitoring module configured to continuously monitor neural activity related to the user's trigeminal nerve afferent pathway by stimulating and recording trigeminal nerve somatosensory evoked potentials, and output neural activation signals representing the degree of neural activation; a transcranial alternating current stimulation module including stimulation electrodes and a controllable current source; the transcranial alternating current stimulation module is configured to apply transcranial alternating current stimulation to the user according to instructions, and to issue diagnostic electrical pulses to stimulate trigeminal nerve somatosensory evoked potentials according to instructions; and a control unit, connected to the blood pressure monitoring module, the... The trigeminal nerve activation monitoring module and the transcranial alternating current stimulation module are communicatively connected and configured to: continuously acquire the user's blood pressure data through the blood pressure monitoring module and run an intervention program when the blood pressure data is determined to be hypertension data; wherein, the intervention program includes: instructing the transcranial alternating current stimulation module to apply therapeutic transcranial alternating current stimulation to the user; during the application of therapeutic transcranial alternating current stimulation, instructing the trigeminal nerve activation monitoring module to acquire nerve activation signals and use the nerve activation signals as neural feedback; and instructing the blood pressure monitoring module to continuously acquire the user's blood pressure data and use the blood pressure data as physiological feedback; and adjusting the stimulation parameters of the transcranial alternating current stimulation module according to the neural feedback and the physiological feedback.

[0006] The closed-loop neuromodulation system in this application regulates the stimulation parameters of transcranial alternating current stimulation (tACS) through dual feedback of neural and physiological feedback, ensuring that the therapeutic stimulation is truly and effectively applied to the intended neural target. Through closed-loop control, the system can "titrate" the most suitable tACS dose for each user in real time, thereby maximizing the therapeutic effect and changing the previous "blind" stimulation mode based on fixed protocols, thus improving the safety of treatment.

[0007] In one embodiment, adjusting the stimulation parameters of the transcranial alternating current stimulation module based on the neural feedback and the physiological feedback includes: adjusting the stimulation intensity of the transcranial alternating current stimulation module based on the neural feedback; and adjusting the stimulation frequency of the transcranial alternating current stimulation module based on the physiological feedback.

[0008] In one embodiment, adjusting the stimulation intensity of the transcranial alternating current stimulation module based on the neural feedback includes:

[0009] A preset target activation range is defined. If the neural feedback is below the target activation range, the control unit instructs the transcranial alternating current stimulation module to gradually increase the stimulation intensity until the neural feedback enters the target activation range.

[0010] If the neural feedback is above the target activation range, the control unit instructs the transcranial alternating current stimulation module to gradually reduce the stimulation intensity until the neural feedback enters the target activation range; and

[0011] If the neural feedback is within the target activation range, the current stimulation intensity is maintained.

[0012] In one embodiment, adjusting the stimulation frequency of the transcranial alternating current stimulation module based on the physiological feedback includes: setting a preset blood pressure drop rate threshold; obtaining the blood pressure drop rate based on the physiological feedback; and if the blood pressure drop rate is less than the blood pressure drop rate threshold when the neural feedback is within the target activation range, the control unit initiates a frequency exploration program to adjust the stimulation frequency and find the stimulation frequency that can elicit the fastest blood pressure response.

[0013] In one embodiment, the instruction for the trigeminal nerve activation monitoring module to acquire neural activation signals includes acquiring them using a real-time TSEP analysis algorithm; the modeling process of the real-time TSEP analysis algorithm is as follows:

[0014] The trigeminal nerve activation monitoring module synchronously records EEG data within a time window of the diagnostic electrical pulse to obtain a diagnostic stimulus data block;

[0015] The subsections in the diagnostic stimulus data block are obtained, and the low-frequency drift noise and high-frequency noise in the subsections are filtered by a digital bandpass filter to obtain the diagnostic stimulus denoised data block;

[0016] A preset amplitude threshold is used to obtain the peak and valley values ​​of the diagnostic stimulus denoising data block. Based on the peak and valley values ​​and the amplitude threshold, artifact removal data blocks are obtained by conditional judgment.

[0017] The enhanced waveform is obtained by processing the aforementioned artifact removal data block through an adaptive noise cancellation structure.

[0018] Based on the enhanced waveform, a multidimensional feature vector is obtained through real-time feature extraction using an improved robust peak detection algorithm; and

[0019] The neural activation signal is obtained by weighted summation based on the multidimensional feature vector.

[0020] In one embodiment, the adaptive noise cancellation structure processing procedure is as follows:

[0021] A single data block is obtained from the artifact removal data block, the desired trigeminal nerve somatosensory evoked potential signal and the background noise signal are obtained, and they are linearly superimposed to obtain a modeled data block as the main input;

[0022] The baseline trigeminal somatosensory evoked potential signal template waveform was obtained by measuring the trigeminal somatosensory evoked potential signal using the traditional signal superposition and averaging method, and used as a reference input.

[0023] Obtain the filter, adjust the filter's weight vector using a normalized least mean square algorithm, and output an adaptive filter; and

[0024] The artifact removal data block is passed through the adaptive filter to obtain an enhanced data block, and the enhanced data block is averaged to obtain the enhanced waveform.

[0025] In one embodiment, the real-time feature extraction process of the improved robust peak detection algorithm is as follows:

[0026] A preset time window is used to calculate the signal moving average and the signal moving standard deviation within the time window.

[0027] The data threshold is obtained by adding three times the standard deviation of the signal movement to the moving average of the signal.

[0028] Data points of the enhanced waveform are obtained. If the data point is greater than the data threshold, the data point is marked as a potential peak.

[0029] Based on the potential peak value, the point with the largest amplitude is selected as the characteristic peak within the time window; and

[0030] Based on the characteristic peaks, the latency and amplitude are identified, and the multidimensional feature vector is constructed.

[0031] In one embodiment, the blood pressure monitoring module includes a hybrid sensor, a noise amplifier, a filter, an analog-to-digital converter, a microprocessor, and an initial blood pressure estimation model; the hybrid sensor includes, but is not limited to, an ECG electrode and a PPG sensor.

[0032] In one embodiment, the process by which the blood pressure monitoring module acquires blood pressure data is as follows:

[0033] The original signal is acquired through the hybrid sensor; based on the original signal, the ECG waveform and PPG waveform are obtained by amplification, filtering, and digital extraction through the noise amplifier, the filter, and the analog-to-digital converter.

[0034] The cardiac contraction delay time is obtained by using the microprocessor to identify the R-wave peak point of the ECG waveform and the characteristic point of the PPG waveform, and calculating the time difference between the R-wave peak point and the characteristic point to obtain the pulse arrival time. The pure pulse conduction time is obtained by subtracting the pulse arrival time from the cardiac contraction delay time.

[0035] The blood pressure data is obtained based on the pure pulse conduction time using the initial blood pressure estimation model.

[0036] In one embodiment, the closed-loop neuromodulation method further includes: generating a diagnostic adjustment pulse through the transcranial alternating current stimulation module based on the stimulation parameters; stimulating the neural activation adjustment signal of the trigeminal nerve activation monitoring module through the diagnostic adjustment pulse; inputting the neural activation adjustment signal as a correction factor into the blood pressure monitoring module to correct the initial blood pressure estimation model to obtain a corrected blood pressure estimation model; and obtaining corrected blood pressure data through the corrected blood pressure estimation model based on the pure pulse conduction time, wherein the corrected blood pressure data includes systolic blood pressure estimation data and diastolic blood pressure estimation data.

[0037] In one embodiment, the blood pressure monitoring module, the trigeminal nerve activation monitoring module, the control unit, and the transcranial AC stimulation module are integrated into one or more wearable components.

[0038] In one embodiment, the wearable component includes a wristband device and a headband device.

[0039] This application introduces neural activation adjustment signals as a real-time proxy indicator for the main source of model drift. The control unit uses this indicator to continuously correct the initial blood pressure estimation model online, thereby significantly reducing model drift and reducing or even eliminating the reliance on frequent manual cuff calibration, truly achieving convenient and reliable continuous blood pressure monitoring. Attached Figure Description

[0040] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0041] Figure 1 This is a structural block diagram of the closed-loop neural modulation system provided in the embodiments of this application;

[0042] Figure 2 A block diagram illustrating the working principle of the closed-loop neural modulation system provided in this application embodiment;

[0043] Figure 3 A schematic diagram of the system provided in this application being worn by a user;

[0044] Figure 4 A flowchart of the closed-loop control method provided in the embodiments of this application;

[0045] Figure 5 The closed-loop neuromodulation system of this application is used to record the curves of changes in the patient's blood pressure, trigeminal nerve activation signal, and stimulation parameters over time when a hypertensive patient undergoes transcranial alternating current stimulation. Detailed Implementation

[0046] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0047] Please see Figure 1 and Figure 2 A closed-loop neuromodulation system for hypertension intervention, comprising a blood pressure monitoring module 100, a trigeminal nerve activation monitoring module 200, a transcranial alternating current stimulation module 400, and a control unit 300.

[0048] The blood pressure monitoring module 100 is configured to continuously monitor the user's blood pressure and output blood pressure data representing the blood pressure level.

[0049] The trigeminal nerve activation monitoring module 200 is configured to continuously monitor neural activity related to the user's trigeminal nerve afferent pathway by stimulating and recording trigeminal somatosensory evoked potentials (TSEPs), and output neural activation signals representing the degree of neural activation.

[0050] The transcranial alternating current stimulation module 400 includes stimulation electrodes and a controllable current source. The transcranial alternating current stimulation module is configured to apply transcranial alternating current stimulation with specific parameters to a user upon instruction, and to emit diagnostic electrical pulses upon instruction to stimulate trigeminal nerve somatosensory evoked potentials.

[0051] The control unit 300 is communicatively connected to the blood pressure monitoring module 100, the trigeminal nerve activation monitoring module 200, and the transcranial alternating current stimulation module 400. The control unit 300 is configured to continuously acquire the user's blood pressure data through the blood pressure monitoring module 100 and, when the blood pressure data is determined to be hypertension data, run an intervention program.

[0052] The intervention procedure includes:

[0053] The transcranial alternating current stimulation module 400 is instructed to apply therapeutic transcranial alternating current stimulation to the user;

[0054] During the application of therapeutic transcranial alternating current stimulation, the trigeminal nerve activation monitoring module 200 is instructed to acquire nerve activation signals and use these signals as neural feedback; and the blood pressure monitoring module 100 is instructed to continuously acquire the user's blood pressure data and use this data as physiological feedback; and

[0055] The stimulation parameters of the transcranial alternating current stimulation module 400 are adjusted based on the neural feedback and the physiological feedback.

[0056] Please continue reading. Figure 3The blood pressure monitoring module 100, the trigeminal nerve activation monitoring module 200, the control unit 300, and the transcranial AC stimulation module 400 are integrated into one or more wearable components, including wristband devices and headband devices, and the modules communicate wirelessly with the devices.

[0057] In one example of a blood pressure monitoring module 100, the blood pressure monitoring module 100 includes a hybrid sensor, a noise amplifier, a filter, an analog-to-digital converter, a microprocessor, and an initial blood pressure estimation model.

[0058] The hybrid sensor includes, but is not limited to, an ECG electrode and a PPG sensor; the ECG electrode can be integrated into the inner and outer metal contacts of the wristband, forming a measurement circuit upon user touch; the PPG sensor is integrated into the inner side of the wristband, close to the skin above the radial artery. This ECG+PPG combination is essential for accurate PTT calculation.

[0059] In this embodiment, the initial blood pressure estimation model includes, but is not limited to, linear regression model, support vector machine model, and neural network model; the parameters in the initial blood pressure estimation model are initially calibrated by performing a one-time synchronous measurement with a standard cuff blood pressure monitor when the device is used for the first time.

[0060] Specifically, the process by which the blood pressure monitoring module 100 acquires blood pressure data is as follows:

[0061] The original signal is acquired through the hybrid sensor; based on the original signal, the ECG waveform and PPG waveform are obtained by amplification, filtering, and digital extraction through the noise amplifier, the filter, and the analog-to-digital converter.

[0062] The heart contraction delay time is obtained by the microprocessor identifying the R-wave peak point of the ECG waveform and the feature point of the PPG waveform, and calculating the time difference between the R-wave peak point and the feature point to obtain the pulse arrival time. The pure pulse conduction time is obtained by subtracting the pulse arrival time from the heart contraction delay time.

[0063] The blood pressure data is obtained based on the pure pulse conduction time using the initial blood pressure estimation model.

[0064] In this embodiment, the characteristic points of the PPG waveform include, but are not limited to, the peak point of the contraction phase and the foot point of the waveform.

[0065] Continue reading Figure 4Specifically, adjusting the stimulation parameters of the transcranial alternating current stimulation module 400 based on the neural feedback and the physiological feedback includes: adjusting the stimulation intensity of the transcranial alternating current stimulation module 400 based on the neural feedback; and adjusting the stimulation frequency of the transcranial alternating current stimulation module 400 based on the physiological feedback.

[0066] In this embodiment, adjusting the stimulation intensity of the transcranial alternating current stimulation module 400 based on the aforementioned neural feedback includes:

[0067] A preset target activation range is defined. If the neural feedback is below the target activation range, the control unit 300 instructs the transcranial alternating current stimulation module 400 to gradually increase the stimulation intensity until the neural feedback enters the target activation range.

[0068] If the neural feedback is higher than the target activation range, the control unit 300 instructs the transcranial alternating current stimulation module 400 to gradually reduce the stimulation intensity until the neural feedback enters the target activation range;

[0069] If the neural feedback is within the target activation range, the current stimulation intensity is maintained; this process ensures the "targetedness" and "effectiveness" of the treatment and achieves closed-loop control of the stimulation intensity.

[0070] In this embodiment, the control unit 300 instructs the transcranial alternating current stimulation module 400 to gradually increase the stimulation intensity by increasing Delt every 30 seconds. I =0.5mA.

[0071] Specifically, adjusting the stimulation frequency of the transcranial alternating current stimulation module 400 based on the physiological feedback includes: setting a preset blood pressure decrease rate threshold; obtaining the blood pressure decrease rate based on the physiological feedback; and if the blood pressure decrease rate is less than the blood pressure decrease rate threshold when the neural feedback is within the target activation range, the control unit 300 initiates a frequency exploration program to adjust the stimulation frequency and find the stimulation frequency that can elicit the fastest blood pressure response.

[0072] In this embodiment, Table 1 shows the tACS stimulation parameter specifications. The stimulation parameters are adjusted according to Table 1. Hypertension intervention is performed by adjusting the frequency, intensity, waveform, and duration of a single intervention. The most suitable tACS dose is "titrated" for each user in real time, thereby maximizing the therapeutic effect while improving the safety of the treatment.

[0073] Table 1

[0074]

[0075] In this embodiment, the trigeminal nerve activation monitoring module 200 records the nerve activation signal via dry EEG electrodes; optionally, the dry EEG electrodes are integrated into a lightweight headband; according to clinical practice of TSEP recording, the recording electrodes are preferably placed at positions C5, C6, and Fz in the international 10-20 system, and the ground electrode is placed in the masseter muscle of the mandible or the shoulder; the electrode material is preferably Ag / AgCl to ensure signal quality, and the inter-electrode impedance should be maintained at Zimp < 5kΩ.

[0076] Specifically, the modeling process of the real-time TSEP analysis algorithm is as follows:

[0077] Diagnostic electrical pulses are acquired, and the trigeminal nerve activation monitoring module 200 simultaneously records EEG data within a time window of the diagnostic electrical pulses to obtain diagnostic stimulus data blocks;

[0078] The subsections in the diagnostic stimulus data block are obtained, and the low-frequency drift noise and high-frequency noise in the subsections are filtered by a digital bandpass filter to obtain the diagnostic stimulus denoised data block;

[0079] A preset amplitude threshold is used to obtain the peak and valley values ​​of the diagnostic stimulus denoising data block. Based on the peak and valley values ​​and the amplitude threshold, artifact removal data blocks are obtained by conditional judgment.

[0080] The mathematical expression for the conditional judgment is:

[0081] max(f{e i (t)})-min(f{e i (t)})>V artifact ,

[0082] Where, maxf{e i (t)} represents the peak value, minf{e i V(t)} represents the valley value, V artifact The amplitude threshold is defined as follows.

[0083] The enhanced waveform is obtained by processing the aforementioned artifact removal data block through an adaptive noise cancellation structure.

[0084] Based on the enhanced waveform, a multidimensional feature vector is obtained by real-time feature extraction using an improved robust peak detection algorithm.

[0085] The neural activation signal is obtained by weighted summation based on the multidimensional feature vector.

[0086] The expression for the weighted summation calculation is:

[0087]

[0088] Wherein, NAS is the neural activation signal, and W A W L1 W L2 A is the weighting coefficient. P19 L N27 During the incubation period, L P19 Let L be the amplitude and Z be the normalization factor; the logic of the expression is: L P19 The larger the amplitude, the more A P19 and L N27 The shorter the latency period, the faster the trigeminal nerve pathway conducts and the higher the degree of activation, and therefore the higher the NAS score.

[0089] In this embodiment, the acquisition of the diagnostic electrical pulses includes: the control unit 300 instructing the transcranial alternating current stimulation module 400 to emit a short series (M=20 times) of diagnostic electrical pulses during the intervals of therapeutic transcranial alternating current stimulation, with a stimulation frequency of f. stim .

[0090] Optionally, the digital bandpass filter is a fourth-order Butterworth filter, and the passband range of the fourth-order Butterworth filter is [f low ,f high ], where f low =30Hz, f high =750Hz, to filter out low-frequency drift and high-frequency noise.

[0091] In this embodiment, the function of the trigeminal nerve activation monitoring module 200 is tightly integrated with the transcranial alternating current stimulation module 400. When neural feedback is required, the control unit 300 instructs the transcranial alternating current stimulation module 400 to emit a series of diagnostic electrical pulses to stimulate TSEP through electrodes placed on the face (e.g., the upper and lower lip areas). The parameters of the diagnostic electrical pulses are preferably a series of monophasic rectangular pulses, characterized by a pulse width delta. stim Current intensity I stim ∈[1,2]mA, repetition frequency f stim ∈[3,4]Hz is defined.

[0092] In this embodiment, the transcranial alternating current stimulation module 400 includes, but is not limited to, two pairs of electrodes; one pair is used for therapeutic tACS, preferably hydrogel electrodes or Ag / AgCl electrodes with good conductivity and biocompatibility, placed in the forehead and bilateral or unilateral mastoid regions. The other pair is used for diagnostic TSEP stimulation, and can be patch electrodes placed in the distribution area of ​​the trigeminal nerve endings, such as the upper and lower lips. The current source of the transcranial alternating current stimulation module is a programmable constant current source precisely controlled by the control unit 300. The constant current source can generate stable and accurate sinusoidal alternating current for treatment, and can also generate high-precision, short-duration monophasic rectangular pulses for TSEP diagnosis according to instructions. The transcranial alternating current stimulation module 400 can adjust the tACS stimulation parameters within a wide treatment range and integrates hardware safety circuitry, including but not limited to: independent output and return current monitors, an error detector composed of hardware comparators, an independent hardware current circuit breaker, and a latching error detector.

[0093] Specifically, tACS is transcranial alternating current stimulation; TSEP is trigeminal nerve somatosensory evoked potential.

[0094] Specifically, the adaptive noise cancellation structure processing procedure is as follows:

[0095] A single data block is obtained from the artifact removal data block. The expected trigeminal nerve somatosensory evoked potential signal and the background noise signal are obtained and linearly superimposed to obtain the modeled data block as the main input.

[0096] The baseline trigeminal somatosensory evoked potential signal template waveform was obtained by measuring the trigeminal somatosensory evoked potential signal using the traditional signal superposition and averaging method, and used as a reference input.

[0097] Obtain the filter, adjust the weight vector of the filter using the normalized least mean square algorithm, and output an adaptive filter.

[0098] The artifact removal data block is passed through the adaptive filter to obtain an enhanced data block, and the enhanced data block is averaged to obtain the enhanced waveform.

[0099] Specifically, the real-time feature extraction process of the improved robust peak detection algorithm is as follows:

[0100] A preset time window is used to calculate the signal moving average and the signal moving standard deviation within the time window.

[0101] The data threshold is obtained by adding three times the standard deviation of the signal movement to the moving average of the signal.

[0102] Data points of the enhanced waveform are acquired. If the data point is greater than the data threshold, the data point is marked as a potential peak.

[0103] Based on the potential peak value, the point with the largest amplitude is selected as the characteristic peak within the time window;

[0104] Based on the characteristic peaks, the latency and amplitude are identified, and the multidimensional feature vector is constructed.

[0105] In this embodiment, the characteristic peaks include N13, P19, and N27; the latency of the characteristic peaks is LN13, LP19, and LN27; the amplitude of the characteristic peaks is AN13, AP19, and AN27; and the multidimensional feature vector is represented as: F = [LN13, AN13, LP19, AP19, LN27, AN27] T .

[0106] Specifically, the closed-loop neural modulation method further includes: generating diagnostic adjustment pulses through the transcranial alternating current stimulation module 400 based on the stimulation parameters; stimulating the neural activation adjustment signal of the trigeminal nerve activation monitoring module 200 through the diagnostic adjustment pulses; inputting the neural activation adjustment signal as a correction factor into the blood pressure monitoring module 100 to correct the initial blood pressure estimation model to obtain a corrected blood pressure estimation model; obtaining corrected blood pressure data through the corrected blood pressure estimation model based on the pure pulse conduction time, wherein the corrected blood pressure data includes systolic blood pressure estimation data and diastolic blood pressure estimation data;

[0107] The corrected blood pressure estimation model is expressed as follows:

[0108] BPcorrected(t)=(α·τPTT(t)+β)+f(NAS(t)),

[0109] Wherein, BPcorrected(t) represents the corrected blood pressure data, τPTT(t) is the real-time measured pure pulse conduction time, NAS(t) is the real-time neural activation adjustment signal, f is the correction function, and α and β are weighting coefficients. When the neural activation signal is enhanced (e.g., representing increased sympathetic nerve activity), the correction function outputs a positive value, and vice versa. In this way, the system can automatically compensate for relationship drift caused by changes in the state of the autonomic nervous system in real time, thereby continuously outputting more accurate blood pressure values.

[0110] In this embodiment, the transcranial alternating current stimulation module 400 can also be replaced with a transcranial direct current stimulation module and a transcranial random noise stimulation module, and the stimulation parameter instructions in the control unit 300 can be adjusted to be applicable to transcranial direct current stimulation (tDCS) or transcranial random noise stimulation (tRNS).

[0111] In this embodiment, the control unit 300 can also implement a reinforcement learning algorithm. Within the reinforcement learning framework: the state space S is a multi-dimensional vector s consisting of the current blood pressure value, blood pressure change trend, and neural activation signal values. t ∈S; Action space A: Stimulus parameter combination (frequency, intensity) of tACS a t ∈A; reward function R(s) t ,a t ,s t+1 ): Defined based on the magnitude, speed, and duration of the decrease in blood pressure within the normal range; through continuous "trial and error" and learning, the RL agent (running in the control unit) can autonomously learn an optimal strategy designed to maximize the expected cumulative discount reward.

[0112] In this embodiment, the system can also be equipped with a companion smartphone application for visualization; it connects to the device via Bluetooth to visually display the user's long-term blood pressure trend, the frequency and duration of hypertensive events, and the record of tACS intervention, and allows the user to record life events (such as diet, exercise, emotional stress) and subjective feedback for the purpose of optimizing the control algorithm.

[0113] The following is combined Figure 5 The following example illustrates the working process of the closed-loop neuromodulation system in this case, using a hypertensive patient undergoing transcranial alternating current stimulation.

[0114] The system first continuously monitors the patient's blood pressure. Initially, the blood pressure is high, exceeding a preset hypertension threshold. At this point, the system does not intervene. Once the system detects persistently high blood pressure, it automatically initiates an intervention program, applying an initial intensity of transcranial alternating current stimulation (tACS). This stimulation causes the blood pressure to slowly decrease. During the intervention, the system periodically measures a key neural activation signal. The system compares the actual value of this signal to a preset optimal target range. During monitoring, the first measurement shows the signal value is below the target range, indicating that the initial stimulation intensity is insufficient to most effectively activate the relevant neural pathways. Therefore, the control unit, based on this neural signal feedback, decides to increase the tACS stimulation intensity. This decision is based on the neural signal, not the current blood pressure value. After increasing the stimulation intensity, the blood pressure decreases more rapidly. Subsequent neural signal measurements show that the value has entered the target range, and the system maintains this intensity to continue the intervention. Finally, when the system detects that the blood pressure has decreased to the normal range and remained stable, the intervention is considered successful. Subsequently, the system smoothly reduces the stimulation intensity to zero and returns to the initial continuous monitoring state.

[0115] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope. The scope of protection of the present invention is defined by the appended claims, specification, and their equivalents.

Claims

1. A closed-loop neural modulation system for hypertension intervention, characterized in that, The system includes: A blood pressure monitoring module (100) is configured to continuously monitor the user's blood pressure and output blood pressure data representing the blood pressure level for acquiring the user's blood pressure data; The trigeminal nerve activation monitoring module (200) is configured to continuously monitor neural activity related to the user's trigeminal nerve afferent pathway by stimulating and recording trigeminal nerve somatosensory evoked potentials, and output neural activation signals representing the degree of neural activation. A transcranial alternating current stimulation module (400) includes stimulation electrodes and a controllable current source; the transcranial alternating current stimulation module is configured to apply transcranial alternating current stimulation to a user according to instructions, and to emit diagnostic electrical pulses according to instructions to stimulate trigeminal nerve somatosensory evoked potentials; and A control unit (300) is communicatively connected to the blood pressure monitoring module (100), the trigeminal nerve activation monitoring module (200), and the transcranial alternating current stimulation module (400), and is configured to: continuously acquire the user's blood pressure data through the blood pressure monitoring module (100) and run an intervention program when the blood pressure data is determined to be hypertension data; wherein the intervention program includes: The transcranial alternating current stimulation module (400) is instructed to apply therapeutic transcranial alternating current stimulation to the user; During the application of therapeutic transcranial alternating current stimulation, the trigeminal nerve activation monitoring module (200) is instructed to acquire nerve activation signals and use these signals as neural feedback; and the blood pressure monitoring module (100) is instructed to continuously acquire the user's blood pressure data and use this data as physiological feedback; and The stimulation parameters of the transcranial alternating current stimulation module (400) are adjusted according to the neural feedback and the physiological feedback.

2. The system according to claim 1, characterized in that, The method of adjusting the stimulation parameters of the transcranial alternating current stimulation module (400) based on the neural feedback and the physiological feedback includes: adjusting the stimulation intensity of the transcranial alternating current stimulation module (400) based on the neural feedback; and adjusting the stimulation frequency of the transcranial alternating current stimulation module (400) based on the physiological feedback.

3. The system according to claim 2, characterized in that, The adjustment of the stimulation intensity of the transcranial alternating current stimulation module (400) based on the neural feedback includes: If the neural feedback is below the target activation range, the control unit (300) instructs the transcranial AC stimulation module (400) to gradually increase the stimulation intensity until the neural feedback enters the target activation range. If the neural feedback is above the target activation range, the control unit (300) instructs the transcranial alternating current stimulation module (400) to gradually reduce the stimulation intensity until the neural feedback enters the target activation range; and If the neural feedback is within the target activation range, the current stimulation intensity is maintained.

4. The system according to claim 3, characterized in that, The method of adjusting the stimulation frequency of the transcranial alternating current stimulation module (400) based on the physiological feedback includes: setting a blood pressure drop rate threshold, obtaining the blood pressure drop rate based on the physiological feedback; if the blood pressure drop rate is less than the blood pressure drop rate threshold when the neural feedback is within the target activation range, the control unit (300) starts a frequency exploration program to adjust the stimulation frequency and find the stimulation frequency that can cause the fastest blood pressure response.

5. The system according to claim 1, characterized in that, The instruction to the trigeminal nerve activation monitoring module (200) to acquire neural activation signals includes using a real-time TSEP analysis algorithm; the modeling process of the real-time TSEP analysis algorithm is as follows: The trigeminal nerve activation monitoring module (200) synchronously records EEG data within a time window of the diagnostic electrical pulse to obtain a diagnostic stimulus data block; The subsections in the diagnostic stimulus data block are obtained, and the low-frequency drift noise and high-frequency noise in the subsections are filtered by a digital bandpass filter to obtain the diagnostic stimulus denoised data block; A preset amplitude threshold is used to obtain the peak and valley values ​​of the diagnostic stimulus denoising data block. Based on the peak and valley values ​​and the amplitude threshold, artifact removal data blocks are obtained by conditional judgment. The enhanced waveform is obtained by processing the aforementioned artifact removal data block through an adaptive noise cancellation structure. Based on the enhanced waveform, a multidimensional feature vector is obtained by real-time feature extraction using an improved robust peak detection algorithm. as well as The neural activation signal is obtained by weighted summation based on the multidimensional feature vector.

6. The system according to claim 5, characterized in that, The adaptive noise cancellation structure processing procedure is as follows: A single data block is obtained from the artifact removal data block, the desired trigeminal nerve somatosensory evoked potential signal and the background noise signal are obtained, and they are linearly superimposed to obtain a modeled data block as the main input; The baseline trigeminal somatosensory evoked potential signal template waveform was obtained by measuring the trigeminal somatosensory evoked potential signal using the traditional signal superposition and averaging method, and used as a reference input. Obtain the filter, adjust the weight vector of the filter using the normalized least mean square algorithm, and output an adaptive filter. as well as The artifact removal data block is passed through the adaptive filter to obtain an enhanced data block, and the enhanced data block is averaged to obtain the enhanced waveform.

7. The system according to claim 5, characterized in that, The real-time feature extraction process of the improved robust peak detection algorithm is as follows: A preset time window is used to calculate the signal moving average and the signal moving standard deviation within the time window. The data threshold is obtained by adding three times the standard deviation of the signal movement to the moving average of the signal. Data points of the enhanced waveform are obtained. If the data point is greater than the data threshold, the data point is marked as a potential peak. Based on the potential peak value, the point with the largest amplitude is selected as the characteristic peak within the time window; as well as Based on the characteristic peaks, the latency and amplitude are identified, and the multidimensional feature vector is constructed.

8. The system according to claim 1, characterized in that, The blood pressure monitoring module (100) includes a hybrid sensor, a noise amplifier, a filter, an analog-to-digital converter, a microprocessor, and an initial blood pressure estimation model; the hybrid sensor includes, but is not limited to, an ECG electrode and a PPG sensor.

9. The system according to claim 8, characterized in that, The process by which the blood pressure monitoring module (100) acquires blood pressure data is as follows: The original signal is acquired through the hybrid sensor; based on the original signal, the ECG waveform and PPG waveform are obtained by amplification, filtering, and digital extraction through the noise amplifier, the filter, and the analog-to-digital converter. The cardiac contraction delay time is obtained by using the microprocessor to identify the R-wave peak point of the ECG waveform and the characteristic point of the PPG waveform, and calculating the time difference between the R-wave peak point and the characteristic point to obtain the pulse arrival time. The pure pulse conduction time is obtained by subtracting the pulse arrival time from the cardiac contraction delay time. The blood pressure data is obtained based on the pure pulse conduction time using the initial blood pressure estimation model.

10. The system according to claim 9, characterized in that, The closed-loop neural modulation method further includes: generating a diagnostic adjustment pulse through the transcranial alternating current stimulation module (400) based on the stimulation parameters; stimulating the neural activation adjustment signal of the trigeminal nerve activation monitoring module (200) through the diagnostic adjustment pulse; inputting the neural activation adjustment signal as a correction factor into the blood pressure monitoring module (100) to correct the initial blood pressure estimation model to obtain a corrected blood pressure estimation model; obtaining corrected blood pressure data through the corrected blood pressure estimation model based on the pure pulse conduction time, wherein the corrected blood pressure data includes systolic blood pressure estimation data and diastolic blood pressure estimation data.

11. The system according to claim 1, characterized in that, The blood pressure monitoring module (100), the trigeminal nerve activation monitoring module (200), the control unit (300), and the transcranial AC stimulation module (400) are integrated into one or more wearable components.

12. The system according to claim 11, characterized in that, The wearable components include wristband devices and head-mounted devices.

Images