A transcutaneous electrical stimulation method for autonomic nervous balance modulation

Abstract

The application provides a transcutaneous electrical stimulation method for autonomic nervous balance regulation, which comprises the following steps: obtaining a regulation target, and determining at least one stimulation point according to the regulation target; in each preset time window, the control vector of the time window is calculated by the following steps: calculating a physiological characteristic value according to a multimodal physiological signal of a user; constructing an actual state vector according to the physiological characteristic value; determining a state error according to the actual state vector; updating an online adjustment vector according to the state error; finally, generating the control vector of the current time window; and outputting a corresponding stimulation waveform on the stimulation point. The method constructs a complete regulation link of physiological perception-state estimation-error calculation-parameter update-closed loop output, and the closed loop mechanism enables the control vector to be adaptively adjusted according to the change of the state of the user, greatly improving the individual adaptability, effect stability and use comfort of the transcutaneous electrical stimulation.

Description

Technical Field

[0001] This invention relates to the field of neuromodulation technology, and in particular to a transcutaneous electrical stimulation method for autonomic nervous system balance regulation. Background Technology

[0002] The autonomic nervous system plays a crucial role in cardiovascular regulation, respiratory regulation, stress response, sleep preparation, emotional homeostasis, and the maintenance of visceral function. Autonomic nervous system imbalances are commonly associated with fatigue, anxiety, sleep disorders, chronic stress, pain sensitization, and various functional disorders. Therefore, interventions and regulation of autonomic nervous system function have become important research directions in the fields of neuromodulation, rehabilitation engineering, and digital healthcare.

[0003] Transcutaneous electrical stimulation (TENS) is a non-invasive technique that applies electrical pulses to nerves or tissues via electrodes on the skin to achieve neuromodulation. Due to its advantages such as high safety, ease of operation, and relatively low cost, it has been widely used in many scenarios, including sleep assistance, relaxation training, stress intervention, rehabilitation therapy, and neurofunctional regulation. Particularly in areas such as the ear (including the periauricular region), neck, upper limbs, and chest and abdomen, by setting appropriate electrical stimulation parameters, it can regulate pathways such as the vagus nerve, thereby affecting the sympathetic-parasympathetic balance and promoting the body's transition to a relaxed or recovery state. It can also assist in the recovery of organ function.

[0004] Current transcutaneous electrical stimulation (TES) protocols in clinical and daily use mostly employ open-loop control, meaning they output fixed parameters such as current, frequency, and pulse width according to a preset program, lacking the ability to perceive and dynamically adjust to changes in the user's real-time physiological state. However, the autonomic nervous system's state changes dynamically during stimulation and is easily affected by various factors such as respiration, body movement, emotional fluctuations, fatigue levels, and the external environment. When the stimulation system cannot adjust parameters based on the user's real-time physiological feedback, it is prone to problems such as insufficient stimulation (insignificant effect) or excessive stimulation (causing discomfort or even pain), leading to large fluctuations in treatment effectiveness and a poor user experience. Furthermore, although some protocols incorporate physiological monitoring, they often rely solely on a single indicator such as heart rate or heart rate variability, which is insufficient to comprehensively and accurately reflect the multidimensional autonomic nervous system state, including sympathetic and parasympathetic activity, recovery level, and stress level, resulting in state estimation bias and consequently affecting the accuracy of regulation.

[0005] Therefore, existing technologies suffer from at least the following drawbacks: a lack of closed-loop control mechanisms based on real-time feedback of multimodal physiological signals, and an inability to dynamically optimize stimulation parameters according to the user's real-time physiological state, resulting in poor individual adaptability, insufficient stability of effects, and a need to improve the user experience of transcutaneous electrical stimulation. There is an urgent need for a closed-loop control method that can integrate multimodal physiological signals, estimate the autonomic nervous system state in real time, and automatically adjust stimulation parameters. Summary of the Invention

[0006] In view of the above-mentioned deficiencies of the prior art, the present invention provides a transcutaneous electrical stimulation method for autonomic nerve balance regulation, so as to solve the technical problem of insufficient dynamic adaptability of parameters in closed-loop regulation.

[0007] To achieve the above and other related objectives, the present invention provides a transcutaneous electrical stimulation method for autonomic nervous system balance regulation, comprising: acquiring a regulation target and determining at least one stimulation point based on the regulation target; calculating a control vector for each preset time window according to the following steps: acquiring the user's multimodal physiological signals in the current time window and calculating physiological characteristic values ​​based on the multimodal physiological signals; constructing the user's actual state vector in the current time window based on the physiological characteristic values; determining a state error based on the actual state vector and a preset target state vector corresponding to the regulation target; updating the online adjustment vector of the current time window based on the state error and the online adjustment vector of the previous time window; generating a control vector for the current time window based on a preset basic control vector, the state error, and the online adjustment vector of the current time window; and outputting a corresponding stimulation waveform at the stimulation point based on the control vector of the current time window.

[0008] In one embodiment of the present invention, the multimodal physiological signals include electrocardiogram (ECG) signals or pulse wave signals, skin conductance signals, respiratory signals, and body movement signals; the physiological characteristic values ​​include: heart rate (HR) calculated based on the ECG signals or pulse wave signals. k Heart rate variability index RMSSD k The skin conductance level (EDA) calculated based on the aforementioned skin conductance signal. k The respiratory stability index SI calculated based on the respiratory signals k The body motion feature Act calculated based on the body motion signal k .

[0009] In one embodiment of the present invention, the user's actual state vector in the current time window is: ,in: , used to characterize the degree of sympathetic activity; , used to characterize the degree of parasympathetic activity; Used to characterize the degree of recovery; , used to characterize the degree of stress; , used to characterize the degree of sleep readiness; where, a i b i c i d i e i is the preset mapping weight; norm is the normalization function.

[0010] In one embodiment of the present invention, the online adjustment vector of the current time window is: , These are the compensation amounts for stimulus amplitude, modulation frequency, carrier frequency, and pulse width, respectively; the online adjustment vector for the current time window is updated according to the following formula:

[0011] ,

[0012] ,

[0013] ,

[0014] ,

[0015] In the formula:

[0016] , which is the online adjustment vector for the previous time window;

[0017] , is the state error;

[0018] , is the target state vector; μ A ν A μ m ν c μ τ To update the step size.

[0019] In one embodiment of the present invention, generating a control vector for the current time window based on a preset basic control vector, the state error, and the online adjustment vector of the current time window includes: generating an initial control vector for the current time window based on the basic control vector, the state error, and the online adjustment vector of the current time window; calculating the channel interference degree of any two stimulus points in the current time window based on the initial control vector of the current time window; calculating the interference pressure of each stimulus point in the current time window based on the channel interference degree of all stimulus points in the current time window; and correcting the initial control vector using the interference pressure of each stimulus point in the current time window to obtain the control vector for the current time window.

[0020] In one embodiment of the present invention, the initial control vector of the current time window And generate it according to the following formula:

[0021] ,

[0022] ,

[0023] ,

[0024] ,

[0025] ,

[0026] In the formula, , which is the basic control vector; The control gain for the m-th stimulus point; These represent the stimulus amplitude, modulation frequency, carrier frequency, pulse width, and phase in the control vector, respectively; the subscript m indicates the m-th stimulus point, the subscript k indicates the k-th time window, the superscript raw indicates the initial control vector, and the superscript base indicates the base control vector.

[0027] In one embodiment of the present invention, , .

[0028] In one embodiment of the present invention, the channel interference Γ of the m-th stimulation point and the n-th stimulation point within time window k is... mn,k for:

[0029] ,

[0030] In the formula, A max For the maximum permissible stimulus amplitude, σ c σ is a carrier frequency difference scale. m The modulation frequency difference scale; the interference pressure Π of the m-th stimulus point within time window k. m,k for:

[0031] ,

[0032] In the formula, M c This represents the total number of stimulus points.

[0033] In one embodiment of the present invention, outputting a corresponding stimulation waveform at the stimulation point according to the control vector of the current time window includes: calculating a comprehensive safety index according to the control vector of the current time window; determining whether the comprehensive safety index is greater than a set safety threshold; if so, scaling the control vector of the current time window according to the ratio of the comprehensive safety index to obtain a scaled control vector, and outputting a corresponding stimulation waveform at the stimulation point according to the scaled control vector; if not, directly outputting a corresponding stimulation waveform at the stimulation point according to the control vector of the current time window.

[0034] In one embodiment of the present invention, the control vector for the m-th stimulus point within time window k is:

[0035] .

[0036] In one embodiment of the present invention, calculating a comprehensive safety index based on the control vector of the current time window includes: calculating the stimulus amplitude A in the control vector of the current time window. m,k and pulse width τ m,k and the effective electrode contact area A e Calculate the charge density Q using the following formula. mk :

[0037] ;

[0038] Based on the control vector of the current time window Charge density Q m,k and current skin temperature (Temp) k The comprehensive safety index C is calculated using the following formula. safe,k :

[0039] ,

[0040] In the formula, The maximum permissible stimulation amplitude, maximum permissible pulse width, maximum permissible charge density, and maximum permissible skin temperature are defined as follows:

[0041] In one embodiment of the present invention, the expression of the stimulation waveform is as follows:

[0042] ,

[0043] Or:

[0044] ,

[0045] In the formula, p m,k () is the pulse gating function, which takes the value 1 when t falls within the pulse opening interval and takes the value 0 when t falls within the pulse closing interval. (mod) The modulation depth.

[0046] In one embodiment of the present invention, the stimulation method further includes an abnormal stopping step during the stimulation process, the abnormal stopping step including: obtaining a pain score for the current time window. Based on the pain score within the current time window Heart rate (HR) k and skin conductance level EDA k Calculate the abnormal stop value E according to the following formula. stop,k :

[0047] ,

[0048] In the formula, HR lim EDA lim C lim These are the upper limit of safe heart rate, the threshold for abnormal skin conductance, and the upper limit of permissible pain. This is an indicator function that takes the value 1 if the condition is true and 0 otherwise; ∨ represents a logical "OR" operation; it stops when an abnormal value E is encountered. stop,k When the value equals 1, stimulation stops and a safety warning signal is generated.

[0049] In one embodiment of the present invention, the stimulation method further includes a basic control vector update step at the end of stimulation, the basic control vector update step comprising: evaluating the stimulation effect of each time window; and obtaining the control parameters of the time window with the best stimulation effect based on the evaluation results of each time window. Based on the previous base control vector Control parameters for the time window of optimal stimulation effect The updated basic control vector is obtained by the following formula. : In the formula, ρ u This is the update step size for the basic control vector.

[0050] The beneficial effects of this invention are as follows: This invention proposes a transcutaneous electrical stimulation method for autonomic nervous system balance regulation. This method constructs a complete regulatory chain of physiological perception, state estimation, error calculation, parameter update, and closed-loop output. By converting multimodal physiological signals into multidimensional state vectors (sympathetic, parasympathetic, recovery, stress, and sleep preparation), and using state errors to drive the recursive update of the online adjustment vectors, the stimulation parameters are dynamically adapted to the user's real-time physiological changes. This closed-loop mechanism enables the stimulation amplitude, modulation frequency, carrier frequency, pulse width, and phase to adaptively adjust with changes in the user's state, greatly improving the individual adaptability, effect stability, and user comfort of transcutaneous electrical stimulation, and solving the technical problems of one-size-fits-all and delayed response in traditional open-loop schemes. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. The accompanying drawings are incorporated in and constitute a part of this specification, illustrating embodiments consistent with this application, and are used together with the description to explain the principles of this application. Obviously, the drawings described below are merely some embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without creative effort.

[0052] Figure 1 A flowchart illustrating a transcutaneous electrical stimulation method for autonomic nervous system balance regulation, provided as an embodiment of the present invention;

[0053] Figure 2 This is a flowchart of the stimulation point determination process provided in an embodiment of the present invention;

[0054] Figure 3 A flowchart illustrating the calculation of the control vector according to an embodiment of the present invention;

[0055] Figure 4 A detailed flowchart of step S250 provided in an embodiment of the present invention. Detailed Implementation

[0056] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. It should be noted that, unless otherwise specified, the following embodiments and features can be combined with each other. In addition to the specific methods, equipment, and materials used in the embodiments, based on the knowledge of the prior art and the description of the present invention by those skilled in the art, any prior art methods, equipment, and materials similar to or equivalent to the methods, equipment, and materials in the embodiments of the present invention can be used to implement the present invention.

[0057] It should be understood that the terminology used in the embodiments of this invention is for describing specific implementations and not for limiting the scope of protection of this invention. Unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art.

[0058] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In some embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0059] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functions, and operations that may be implemented in the methods and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0060] Please see Figure 1 , Figure 1 A transcutaneous electrical stimulation method for autonomic nerve balance regulation is provided in one embodiment of the present invention, comprising steps S100 to S300.

[0061] Step S100: Obtain the regulation target and determine at least one stimulation point based on the regulation target.

[0062] Please see Figure 2 In a specific embodiment of the present invention, step S100 includes steps S110 to S150.

[0063] Step S110: Obtain the regulatory goal. The regulatory goal refers to the physiological or psychological regulatory state that the user expects to achieve through this transcutaneous electrical stimulation, such as relaxation assistance, stress relief, sleep preparation, recovery support, organ function regulation, etc. Specifically, relaxation assistance tasks aim to reduce stress levels and increase parasympathetic activity; sleep preparation tasks aim to improve sleep readiness and respiratory stability; recovery support tasks focus on improving recovery and reducing sympathetic activation; and organ function regulation tasks aim to dynamically balance and regulate the functions of various organs throughout the body through the synergistic antagonism of the sympathetic and parasympathetic nervous systems: the heart's rate and contractility are finely regulated to adapt to blood supply demands; bronchial diameter changes with respiratory demand; pupil size regulates light intake; vasoconstriction and vasodilation regulate blood pressure and blood flow distribution; peristalsis and secretion in the digestive tract are inhibited during stress and enhanced during rest; the bladder wall and sphincter muscles of the urinary system work together to control urination; the type and amount of secretion from glands (such as sweat glands and salivary glands) are also adjusted; metabolic direction switches between decomposition and synthesis to manage energy; immune system activity is influenced by neural signals; and different stages of sexual function are also dominated by two separate systems. All these regulations work together to serve the body's immediate response to external stimuli and the long-term maintenance of homeostasis. These control targets are not random inputs, but standardized task options preset in the human-computer interaction module. Each target corresponds to a set of preset target state vectors (such as target sympathetic activity, target parasympathetic activity, etc.) and related control templates, thus providing a clear basis for subsequent region selection, point location screening and parameter control.

[0064] Step S120: Based on the control target, determine the test area from at least one preset candidate area. Candidate areas may include, for example, the ear area, neck area, upper limb area, chest and abdomen area, and lower limb area. The system can preset a mapping relationship between the control target and the test area, allowing the test area to be determined based on the control target and this mapping relationship. Alternatively, based on this, a score can be calculated for each preset candidate area, and the final test area can be selected according to the score.

[0065] Step S130: Calculate the score for each candidate stimulus point within the test area. The score is determined based on the impedance characteristics of the candidate stimulus point and its physiological response to the test stimulus. Each test area contains several preset stimulus points, and the number of stimulus points can vary between different test areas. After determining the test area, all stimulus points within all test areas are considered as candidate stimulus points, and the scores for all these candidate stimulus points are calculated. This provides a basis for selecting stimulus points in subsequent steps.

[0066] Specifically, impedance scores can be obtained based on impedance scanning results of candidate stimulation points, functional scores can be obtained based on physiological responses and / or user subjective feedback caused by trial stimulation of candidate stimulation points, and then the scores of candidate stimulation points can be obtained by weighted summation of impedance scores and functional scores.

[0067] Step S140: Based on the scores of multiple candidate stimulus points and the spatial relationship between any two points among the multiple candidate stimulus points, determine the comprehensive score of at least one combination of candidate stimulus points. Specifically, this includes: for candidate stimulus point combination A', calculating its comprehensive score according to the following formula:

[0068] ,

[0069] in, ,

[0070] In the formula, Score is the overall score of candidate stimulus combination A'. i Candidate stimulus point a i The score, ρ is the preset combination penalty coefficient; Ω ij Candidate stimulus point a i and candidate stimulus point a j The degree of mutual interference between them, r i and r j Candidate stimulus points a i and candidate stimulus point a j The coordinates, σ d This is the preset distance attenuation scale.

[0071] Calculate the overall score for each candidate point. iNext, step S400 further considers the scenario of multi-point joint stimulation. In practical applications, sometimes using a single stimulation point has limited effect, while using multiple stimulation points working together may achieve better modulation effects. However, not any combination of two points can achieve the effect of "1+1=2". If two points are too close in space, their electric fields may overlap and interfere with each other, resulting in excessively strong local stimulation or wasted effect. Therefore, this invention innovatively proposes a comprehensive scoring method for candidate stimulation point combinations. This method first sums the scores of all individual points within the combination, representing the basic effect of the combination; then, it subtracts a penalty term based on the distance between the point pairs. This penalty term adopts an exponential decay form, with a large penalty value when the two points are very close and a penalty value approaching zero when the two points are far apart. In this way, the system can automatically avoid spatially close point pairs while pursuing the maximization of the combined effect, thereby selecting an optimal stimulation combination that can cover the effective area without generating internal interference. This significantly improves the efficiency and rationality of multi-channel, multi-point stimulation strategies.

[0072] Step S150: Candidate stimulation point combinations whose comprehensive scores meet preset conditions are identified as recommended stimulation points corresponding to the control target. After calculating the comprehensive scores of all possible candidate stimulation point combinations, step S150 selects the final recommended stimulation point based on preset conditions (e.g., selecting the combination with the highest comprehensive score, or selecting several combinations with scores exceeding a preset threshold). This recommended stimulation point (or combination of points) will be used in subsequent formal transcutaneous electrical stimulation treatment, thereby ensuring that the starting point of treatment is the optimal choice based on objective data and scientific evaluation.

[0073] Understandably, the recommended stimulation point is just one pole in the stimulation circuit. All recommended stimulation points share a reference electrode, and each recommended stimulation point forms a circuit with the reference electrode. The potential of the reference electrode is 0, and its position is preset.

[0074] Please see Figure 3 In step S200, within each preset time window, the control vector for that time window is calculated according to steps S210-S250. The user's physiological state changes in real time at different time periods, such as stress level, relaxation level, and sleep readiness level. By dividing the time into continuous and non-overlapping time windows (e.g., each time window is 30 seconds long), the system can periodically refresh the estimate of the user's current state and dynamically adjust the stimulation parameters accordingly. This allows the stimulation program to closely follow the user's physiological rhythm changes, achieving true adaptive closed-loop control.

[0075] Step S210: Obtain the user's multimodal physiological signals in the current time window, and calculate physiological feature values ​​based on the multimodal physiological signals.

[0076] In a specific embodiment of the present invention, the multimodal physiological signals include electrocardiogram (ECG) signals or pulse wave signals, skin conductance signals, respiratory signals, and body movement signals; the physiological characteristic values ​​include: heart rate (HR) calculated based on the ECG or pulse wave signals. k Heart rate variability index RMSSD k EDA (Electrodermal Analytical Genetics) level calculated based on electrodermal signals k The respiratory stability index (SI) calculated based on respiratory signals. k Body motion features (Act) calculated based on body motion signals k .

[0077] Understandably, the above parameters can be solved by taking a preset duration (e.g., 30 seconds) as a time window, collecting multimodal physiological signals within the window, and processing them into corresponding physiological feature values.

[0078] For electrocardiogram (ECG) signals or pulse wave signals, detect the intervals between adjacent heartbeats within a time window k:

[0079] ,

[0080] In the formula, t n,k RR is the peak heart rate time of the nth heartbeat within the kth time window. n,k The time interval between adjacent heartbeats. The average heart rate (HR) within the k-th time window. k for:

[0081] .

[0082] The heart rate variability index RMSSD at time window k k for:

[0083] ,

[0084] In the formula, N RR,k This represents the number of RR intervals within the k-th time window.

[0085] EDA level of skin conductance within the k-th time window k for:

[0086] ,

[0087] In the formula, g k [n] represents the nth skin electroreception sample value in the kth time window, N k denoted as the number of skin conductance samples in the k-th time window.

[0088] The respiratory stability index SI in the k-th time window k for:

[0089] ,

[0090] In the formula, std(T) resp,k ) represents the magnitude of the respiratory cycle fluctuation, mean(T) resp,k T represents the average respiratory cycle; resp This represents the respiratory cycle sequence within time window k, obtained by filtering the respiratory signal collected by the respiratory acquisition unit, detecting the time of adjacent respiratory peaks, and calculating the period difference; ϵ is a very small positive number to prevent the denominator from being 0; SI k The closer to 1, the more stable the breathing.

[0091] The body movement characteristics Act within the k-th time window k for:

[0092] ,

[0093] In the formula, The first Triaxial acceleration within a time window; N a,k This represents the number of acceleration samples within the window.

[0094] Step S210 above transforms raw biological signals into quantitative indicators with clear physiological significance by extracting features from multimodal physiological signals (ECG / pulse wave, skin conductance, respiration, and body movement) within a time window. For example, heart rate (HR). k Reflects the rhythm of heart activity, heart rate variability RMSSD k It is one of the gold standards for assessing parasympathetic nerve activity; EDA (electrodermal conductance level) k It is a sensitive indicator for measuring sympathetic arousal and stress levels; the respiratory stability index (SI) k This effectively reflects the stability of the breathing rhythm and is closely related to the relaxed state; body movement characteristics Act k It can identify users' limb movements to eliminate interference from motion artifacts on physiological signals. This multimodal feature extraction method provides a rich and reliable data foundation for subsequent accurate estimation of users' autonomic nervous system states, avoiding the one-sidedness of judgment based on a single indicator.

[0095] Step S220: Construct the user's actual state vector in the current time window based on physiological characteristic values.

[0096] In a specific embodiment of the present invention, the user's actual state vector in the current time window is:

[0097] ,

[0098] in:

[0099] , used to characterize the degree of sympathetic activity;

[0100] , used to characterize the degree of parasympathetic activity;

[0101] Used to characterize the degree of recovery;

[0102] , used to characterize the degree of stress;

[0103] , used to characterize the degree of sleep readiness;

[0104] In the formula, a i b i c i d i e i is the preset mapping weight; norm is the normalization function that maps the result to 0~1.

[0105] The introduction of a multidimensional state vector is one of the core innovations of this invention. Traditional transcutaneous electrical stimulation (TES) protocols typically only consider single indicators such as heart rate or skin conductance, which is insufficient to comprehensively describe the complex autonomic nervous system state of the human body. This method creatively integrates multiple physiological characteristics to construct a five-dimensional state vector, including the degree of sympathetic activity. (Reflects the level of stress) and the level of parasympathetic activity. (Reflecting the degree of relaxation) and recovery (Reflecting the body's self-repair ability), stress level (Reflecting physiological load under external pressure) and sleep readiness level (This reflects whether the body has entered a state suitable for falling asleep). These five components are not simply listed, but rather are a linearly weighted combination of information from different dimensions such as heart rate, heart rate variability, skin conductance, respiration, and body movement, based on physiological principles. The benefits of introducing this multi-dimensional state vector are obvious: it provides the control system with a snapshot of the body's actual physiological state, allowing subsequent control decisions (such as increasing or decreasing stimulus intensity) to be based on more comprehensive and accurate information, rather than relying on a single indicator's biased judgment. This greatly improves the accuracy and intelligence of closed-loop regulation.

[0106] Step S230: Determine the state error based on the actual state vector and the preset target state vector corresponding to the control target. The target state vector corresponding to the control target can be denoted as:

[0107] ,

[0108] State error These three parameters are all vectors, representing the state error. It also consists of five components, which can be written as:

[0109] ,

[0110] in,

[0111] ;

[0112] ;

[0113] ;

[0114] ;

[0115] .

[0116] Step S240: Update the online adjustment vector of the current time window based on the state error and the online adjustment vector of the previous time window.

[0117] In a specific embodiment of the present invention, the online adjustment vector of the current time window is:

[0118] ,

[0119] in, These are the compensation amounts for the stimulus amplitude, modulation frequency, carrier frequency, and pulse width, respectively.

[0120] The online adjustment vector for the previous time window is:

[0121] ,

[0122] The online adjustment vector for the current time window is updated according to the following formula:

[0123] ,

[0124] ,

[0125] ,

[0126] ,

[0127] Where: μ A ν A μ m ν c μ τ To update the step size.

[0128] The introduction of online adjustment vectors is another core innovation of this invention. In closed-loop control systems, relying solely on basic control vectors is insufficient to handle complex and ever-changing physiological dynamics. The online adjustment vectors act as fine-tuners. They consist of four components: stimulus amplitude compensation, modulation frequency compensation, carrier frequency compensation, and pulse width compensation, corresponding to dynamic corrections in different dimensions of the stimulus waveform. Their update rule employs a recursive algorithm based on state errors; for example, insufficient recovery (… Increase amplitude compensation when it is positive; excessive stress ( When positive, reduce the amplitude compensation; insufficient parasympathetic activity ( When positive, increase modulation frequency compensation; when sympathetic activity is too high ( Reduce carrier frequency compensation when it is positive; insufficient sleep preparation ( When the value is positive, pulse width compensation is added. The advantage of this design is that it can predictively correct the basic control parameters online based on the real-time deviation between the current state and the target state, making the changes in stimulus parameters smoother and more continuous, and effectively suppressing overshoot. At the same time, the introduction of online adjustment vectors enhances the robustness of the system; even if the basic control vector is not optimal, the system can quickly approximate the ideal stimulus parameters through online adjustment.

[0129] Step S250: Generate the control vector for the current time window based on the preset basic control vector, state error, and online adjustment vector of the current time window. It can be understood that the generated control vector for the current time window applies to the next time window; that is, within the next time window, the control electrode outputs the corresponding waveform according to the control vector of the current time window.

[0130] Please see Figure 4 In a specific embodiment of the present invention, step S250 specifically includes steps S251 to S254.

[0131] Step S251: Generate the initial control vector for the current time window based on the basic control vector, state error, and online adjustment vector of the current time window.

[0132] Within the current time window, besides the control vector In addition, two other control vectors were introduced: the basic control vector. and initial control vector These three control vectors are represented as follows:

[0133] ,

[0134] ,

[0135] ,

[0136] in, These are the stimulus amplitude, modulation frequency, carrier frequency, pulse width, and phase component in the control vector, respectively. In the basic control vector, the modulation frequency... The carrier frequency can be 20~100Hz. It can be 1~10kHz.

[0137] In this invention, three different levels of control vectors are defined. Basic control vector This is the system's starting point or default template, derived from the cumulative updates of historical best stimulus sessions or the initial presets for the task type. It represents a relatively safe and effective parameter benchmark for the user under the current control objectives. Initial Control Vector Based on the basic control vector, the preliminary target parameters are obtained by integrating the state error of the current time window and the online adjustment vector. These parameters reflect the ideal stimulus state the system aims to achieve based on the current physiological feedback. The final control vector... These are the actual output parameters obtained after correcting for inter-channel interference penalties based on the initial control vector, thus resolving the mutual interference problem when multiple stimulus points work together. The stimulus amplitude A and modulation frequency f are selected. (mod) Carrier frequency f (car) The five components—amplitude, pulse width τ, and phase φ—serve as the core parameters of the control vector because they determine the physical characteristics of transcutaneous electrical stimulation (TES) from different dimensions: amplitude determines the stimulation intensity, modulation frequency determines the rate of change in stimulation intensity, carrier frequency determines the depth of action, pulse width determines the duration of energy injection for each pulse, and phase is used to coordinate the timing of multi-point stimulation. These five parameters are independent yet coupled, together forming a flexible and finely adjustable stimulation parameter space, providing ample freedom for achieving precise and personalized closed-loop control.

[0138] The initial control vector is generated according to the following formula:

[0139] ,

[0140] ,

[0141] ,

[0142] ,

[0143] ,

[0144] In the formula, Let be the control gain for the m-th stimulus point; the subscript m indicates the m-th stimulus point (or the number of channels; if the parameters are considered as feature vectors, then each channel corresponds to a stimulus point, therefore, the channels mentioned below are stimulus points), the subscript k indicates the k-th time window, the superscript raw indicates the initial control vector, and the superscript base indicates the base control vector.

[0145] Understandably, without introducing the initial control vector, channel interferometry, and interference pressure, the calculation result of the above formula for the initial control vector can be directly used as the control vector for the current time window. However, after introducing the channel interferometry and interference pressure, the result calculated by the above formula as the initial control vector needs to be corrected according to the interference pressure to obtain the final output control vector.

[0146] Step S251 above demonstrates how to transform abstract state errors into specific stimulus parameter corrections. This is achieved through a preset control gain (such as...). (etc.), the system can quantitatively calculate how much the stimulus amplitude needs to be adjusted at each stimulus point to compensate for insufficient recovery or excessive stress. Online vector adjustment. As a cumulative correction term, the parameters are further smoothed and fine-tuned. This method, which maps state errors to different control components, directly links parameter adjustment with physiological goals.

[0147] Step S252: Based on the initial control vector of the current time window, calculate the channel interferometry between any two stimulus points within the current time window. The channel interferometry Γ between the m-th and n-th stimulus points within time window k. mn,k for:

[0148] ,

[0149] In the formula, A max For the maximum permissible stimulus amplitude, its σ c σ is a carrier frequency difference scale. m Let A be the modulation frequency difference scale. max The first term involved in the calculation is the amplitude overlap term, which indicates that the interference effect is greater when both channels are strong; if one channel is weak, the superposition effect is smaller. There is σ. c The second term involved in the calculation is the carrier proximity term, which indicates that the closer the carrier frequencies of the two channels are, the closer the exponent term is to 1, meaning they are more likely to produce similar outputs at the fast oscillation level. σ c For example, it can be set to around 40Hz; the greater the frequency difference, the smaller this term, and the weaker the interference. There is σ. mThe third term in the residual calculation is the modulation similarity term, which indicates that the closer the modulation rhythms of the two channels are, the more synchronized the slowly changing envelopes of the outputs are, and therefore the easier it is for them to overlap. σ m For example, it can be set to around 5kHz. The last term is the phase coincidence term. When the two channels are almost in phase, this term is close to 1; when the phase difference is about 90 degrees, this term is smaller; using cosine... 2 The reason for using the non-cosine value is to keep the interference degree non-negative and to indicate that the more the phases overlap, the greater the superposition.

[0150] Step S253: Calculate the interference pressure of each stimulus point within the current time window based on the channel interference of all stimulus points within the current time window. The interference pressure Π of the m-th stimulus point within time window k. m,k for:

[0151] ,

[0152] In the formula, M c Γ represents the total number of stimulation points. If two channels are very strong, have similar frequencies, and are in phase synchronization, then Γ mn,k High Γ increases interference pressure; if the channel strength is low, frequency distinction is obvious, and phase is staggered, then Γ mn,k The lower the value, the less the interference pressure.

[0153] Step S254: Using the interference pressure of each stimulus point in the current time window, the initial control vector is corrected to obtain the control vector for the current time window, which can be expressed by the formula:

[0154] ,

[0155] ,

[0156] ,

[0157] ,

[0158] ,

[0159] In the formula, For interference penalty gain. In this embodiment, the reason for introducing interference pressure and modifying the initial control vector based on the interference pressure to obtain the final control vector is explained as follows: (1) If a certain channel has large interference with other channels, the stimulation amplitude A is appropriately reduced. m,k (2) Appropriately increase the modulation frequency and carrier frequency (3) Reduce pulse width τ m,k and adjusting the modulation frequency (4) Adjusting the phase ϕ can reduce the amount of superimposed stimulation per unit time; m,k This staggers the timing of different channels, reducing synchronous overlap. Therefore, inter-channel interference penalty is essentially a multi-channel decoupling mechanism, enabling multiple stimulation points to operate simultaneously without excessive overlap. Adjusting based on the target state vector allows for dynamic adjustment of different frequencies at different stimulation points during operation, ensuring the feasibility of adjustment, unlike the unstable effects of single-parameter stimulation.

[0160] Step S300: Based on the control vector of the current time window, output the corresponding stimulus waveform at the stimulus point. Based on the control vector of the kth time window output in step S200... The corresponding stimulus waveform can be output using the following formula:

[0161] ,

[0162] Or:

[0163] ,

[0164] In the formula, p m,k () is the pulse gating function, which takes the value 1 when t falls within the pulse opening interval and takes the value 0 or -1 when t falls within the pulse closing interval; the subscript m indicates the m-th stimulation point. p m,k If the function does not have The variable indicates that it does not require an initial phase. A (mod) The modulation depth is represented by a value of [value to be filled in]. .

[0165] In a specific embodiment of the present invention, the above step S300 specifically includes: (1) calculating the comprehensive safety index based on the control vector of the current time window; (2) determining whether the comprehensive safety index is greater than the set safety threshold (the set safety threshold is generally 1): if yes, scaling the control vector of the current time window according to the ratio of the comprehensive safety index to obtain the scaled control vector, and outputting the corresponding stimulation waveform at the stimulation point according to the scaled control vector; if no, directly outputting the corresponding stimulation waveform at the stimulation point according to the control vector of the current time window.

[0166] Comprehensive safety index C safe,kThis invention is a master switch designed to ensure user safety. It monitors multiple safety-related parameters, including single-point stimulation amplitude, pulse width, charge density, and current skin temperature, and takes the maximum value of these ratios. The advantage of this design is that it adopts the "weakest link" principle, meaning that if any indicator (such as charge density) approaches or exceeds its safety limit, the overall safety index will be greater than or equal to 1, thereby triggering the safety protection mechanism. This is far more comprehensive and rigorous than monitoring only a single parameter (such as limiting only the current), and can effectively prevent safety risks caused by improper parameter combinations (such as excessively wide pulse width even with a small current, leading to excessive charge density) or local overheating.

[0167] In a specific embodiment of the present invention, the comprehensive safety index is calculated through the following steps.

[0168] First, based on the stimulus amplitude A in the control vector of the current time window. m,k and pulse width τ m,k and the effective electrode contact area A e Calculate the charge density Q using the following formula. m,k :

[0169] .

[0170] Secondly, based on the control vector of the current time window Charge density Q m,k and current skin temperature (Temp) k Calculate the comprehensive safety index C using the following formula. safe,k :

[0171] ,

[0172] In the formula, The maximum permissible stimulation amplitude, maximum permissible pulse width, maximum permissible charge density, and maximum permissible skin temperature are defined as follows:

[0173] In the above formula, the current skin temperature is Temp. k It can be calculated using the following formula:

[0174] ,

[0175] In the formula, T k [n] represents the nth skin temperature sample within the kth time window, N T,k This represents the number of skin temperature samples.

[0176] The specific calculation process of the comprehensive safety index first defines the charge density Q. m,kThis crucial intermediate quantity, charge density (the amount of electricity passing through an electrode per unit area), is a core parameter for assessing the safety of electrical stimulation. Excessively high values ​​can lead to skin burns or severe discomfort. The calculation formula shows that charge density is directly proportional to the stimulation amplitude and pulse width, and inversely proportional to the electrode contact area. Therefore, even with a low stimulation current, if the pulse width is too wide or the electrode area is too small, the charge density may still exceed the limit. The final comprehensive safety index C... safe,k The maximum value among the five relative ratios (amplitude ratio, carrier frequency ratio, pulse width ratio, charge density ratio, and skin temperature ratio) is taken, where each relative ratio is also the maximum relative ratio of multiple stimulation points. When C safe,k When C ≤ 1, it indicates that all monitoring parameters are within the safe range; when C safe,k When the value is greater than 1, a safety scaling mechanism is triggered, proportionally reducing all relevant parameters back to within the safety boundary. This multi-parameter joint safety assessment and processing method provides a solid guarantee for the safe operation of the closed-loop system.

[0177] In a specific embodiment of the present invention, the stimulation method further includes an abnormal stop step in the stimulation process, the abnormal stop step including: (1) obtaining the pain score of the current time window. (2) Pain score based on the current time window Heart rate (HR) k and skin conductance level EDA k Calculate the abnormal stop value E according to the following formula. stop,k :

[0178] ,

[0179] In the formula, HR lim EDA lim C lim These are the upper limit of safe heart rate, the threshold for abnormal skin conductance, and the upper limit of permissible pain. The indicator function takes the value 1 when the condition is true and 0 otherwise; ∨ represents a logical OR operation; (3) When the abnormal stop value E stop,k When the value equals 1, stimulation stops and a safety warning signal is generated.

[0180] The abnormal shutdown procedure is an additional safety safeguard independent of the aforementioned safe scaling mechanism. It focuses on detecting emergency situations that may require immediate shutdown due to physiological abnormalities or user discomfort. Specifically, it monitors three key metrics in real time: heart rate (HR). k ), skin conductance level (EDA) k ) and user pain rating ( When the heart rate exceeds the preset safety limit, it may indicate that the user is overly stressed or has potential cardiovascular risks; when the skin conductance level is abnormally elevated, it may reflect excessive user stress or electrode contact problems; when the user's reported pain score exceeds the allowable limit, it directly indicates that the current stimulation intensity or mode is intolerable. Abnormal stop value E stop,k The formula uses a logical "OR" operation, meaning that if any one of the three conditions is triggered, the system will immediately and unconditionally stop stimulation and issue a warning. This mechanism ensures that even in the worst-case scenario (such as a failure of the safety scaling module or an abnormal parameter combination), the user's safety and comfort experience are still guaranteed.

[0181] In a specific embodiment of the present invention, the stimulation method further includes a basic control vector update step at the end of stimulation. The basic control vector update step includes: (1) evaluating the stimulation effect of each time window; and (2) obtaining the control parameters of the time window with the best stimulation effect based on the evaluation results of each time window. (3) Based on the base control vector before the update Control parameters for the time window of optimal stimulation effect The updated basic control vector is obtained by the following formula. :

[0182] ,

[0183] In the formula, ρ u This is the update step size for the basic control vector.

[0184] The basic control vector update step is crucial for enabling the system's self-evolution and personalized learning. After a complete stimulus session, the system evaluates each time window of the session to identify the control parameters corresponding to the period with the best stimulus effect. Next, the system updates the basic control vector using an exponential moving average (EMA). The advantage of this smooth update strategy is that it leverages the valuable experience gained in the current session to optimize future basic control templates, allowing the system to gradually adapt to the user's personalized needs. Simultaneously, it prevents drastic changes in the basic template due to occasional fluctuations in a single session (such as abnormal user status or external interference), ensuring system stability and long-term convergence. With increased usage, the basic control vector will continuously approach the user's optimal parameter space, thus achieving true "the more you use it, the better it becomes" effectiveness.

[0185] In summary, this invention proposes a transcutaneous electrical stimulation (TES) method for autonomic nervous system balance regulation. By constructing a complete regulatory chain—from physiological perception and state estimation to error calculation, parameter updating, and closed-loop output—it achieves intelligent and personalized TES. First, this method utilizes multimodal physiological signals (ECG / pulse wave, skin conductance, respiration, and body movement) to construct a five-dimensional autonomic nervous system state vector (sympathetic, parasympathetic, recovery, stress, and sleep preparation), transforming abstract physiological indicators into quantifiable control targets. Based on this, the vector is recursively updated online through state error-driven adjustments, and a channel interference penalty mechanism is used to decouple and coordinate multiple stimulation points, achieving dynamic adaptation of stimulation parameters (amplitude, frequency, pulse width, and phase) to real-time physiological changes in the user. Simultaneously, the system integrates a multi-parameter joint safety assessment and an abnormal shutdown mechanism to ensure the safety and reliability of the regulation process. Through efficacy evaluation and smooth updates of the basic control vector after stimulation, the system can learn and optimize individualized templates from each treatment. Compared with traditional open-loop solutions, this invention significantly improves the individual adaptability, effect stability, multi-channel synergistic efficiency, and user comfort of transcutaneous electrical stimulation, providing a precise, safe, and self-learning closed-loop solution for autonomic nervous system modulation scenarios such as sleep preparation, relaxation assistance, stress relief, and organ function recovery.

[0186] It should be noted that the steps of the various methods described above are only for clarity. In practice, they can be combined into one step or some steps can be split into multiple steps. As long as they contain the same logical relationship, they are all within the scope of protection of this application. Adding insignificant modifications or introducing insignificant designs to the algorithm or process, but without changing the core design of the algorithm and process, are also within the scope of protection of this patent.

[0187] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A transcutaneous electrical stimulation method for autonomic nervous system balance regulation, characterized in that, include: Obtain the regulation target, and determine at least one stimulation point based on the regulation target; Within each preset time window, the control vector for that time window is calculated using the following steps: Acquire the user's multimodal physiological signals within the current time window, and calculate physiological feature values ​​based on the multimodal physiological signals; Based on the physiological characteristic values, construct the user's actual state vector in the current time window; The state error is determined based on the actual state vector and the preset target state vector corresponding to the control target; Update the online adjustment vector of the current time window based on the state error and the online adjustment vector of the previous time window; Based on the preset basic control vector, the state error, and the online adjustment vector of the current time window, the control vector of the current time window is generated. Based on the control vector of the current time window, the corresponding stimulation waveform is output at the stimulation point.

2. The transcutaneous electrical stimulation method for autonomic nervous system balance regulation according to claim 1, characterized in that, The multimodal physiological signals include electrocardiogram signals or pulse wave signals, skin conductance signals, respiratory signals, and body movement signals; The physiological characteristic values ​​include: Heart rate (HR) calculated based on the electrocardiogram signal or pulse wave signal k Heart rate variability index RMSSD k ; The skin conductance level (EDA) calculated based on the skin conductance signal is obtained. k ; The respiratory stability index SI calculated based on the respiratory signal k ; The body movement feature Act calculated based on the body movement signal k ; The user's actual state vector in the current time window is: ， in: , used to characterize the degree of sympathetic activity; , used to characterize the degree of parasympathetic activity; , used to characterize the degree of recovery; , used to characterize the degree of stress; , used to characterize the degree of sleep readiness; In the formula, a i b i c i d i e i is the preset mapping weight; norm is the normalization function.

3. The transcutaneous electrical stimulation method for autonomic nervous system balance regulation according to claim 2, characterized in that, The online adjustment vector for the current time window is: ， These are the compensation amounts for the stimulus amplitude, modulation frequency, carrier frequency, and pulse width, respectively. The online adjustment vector for the current time window is updated according to the following formula: ， ， ， ， In the formula: , which is the online adjustment vector for the previous time window; , is the state error; , is the target state vector; μ A ν A μ m ν c μ τ To update the step size.

4. The transcutaneous electrical stimulation method for autonomic nervous system balance regulation according to claim 3, characterized in that, Based on the preset basic control vector, the state error, and the online adjustment vector of the current time window, the control vector for the current time window is generated, including: Based on the basic control vector, the state error, and the online adjustment vector of the current time window, the initial control vector of the current time window is generated. Based on the initial control vector of the current time window, calculate the channel interference degree between any two stimulus points in the current time window; Calculate the interference pressure of each stimulus point in the current time window based on the channel interference of all stimulus points in the current time window; The initial control vector is modified by using the interference pressure of each stimulus point in the current time window to obtain the control vector for the current time window.

5. The transcutaneous electrical stimulation method for autonomic nervous system balance regulation according to claim 4, characterized in that, Initial control vector for the current time window And generate it according to the following formula: ， ， ， ， ， In the formula, , which is the basic control vector. , ; The control gain for the m-th stimulus point; These represent the stimulus amplitude, modulation frequency, carrier frequency, pulse width, and phase in the control vector, respectively; the subscript m indicates the m-th stimulus point, the subscript k indicates the k-th time window, the superscript raw indicates the initial control vector, and the superscript base indicates the base control vector.

6. The transcutaneous electrical stimulation method for autonomic balance regulation according to claim 5, characterized in that, The channel interference Γ between the m-th and n-th stimulation points within time window k mn,k for: ， In the formula, A max For the maximum permissible stimulus amplitude, σ c σ is a carrier frequency difference scale. m The modulation frequency difference scale; The interference pressure Π at the m-th stimulus point within time window k m,k for: ， In the formula, M c This represents the total number of stimulus points.

7. The transcutaneous electrical stimulation method for autonomic nervous system balance regulation according to claim 1, characterized in that, Based on the control vector of the current time window, output the corresponding stimulus waveform at the stimulus point, including: Calculate the comprehensive safety index based on the control vector of the current time window; Determine whether the overall safety index is greater than the set safety threshold: If so, scale the control vector of the current time window according to the ratio of the comprehensive safety index to obtain the scaled control vector, and output the corresponding stimulation waveform at the stimulation point according to the scaled control vector; If not, output the corresponding stimulus waveform at the stimulus point directly according to the control vector of the current time window.

8. The transcutaneous electrical stimulation method for autonomic balance regulation according to claim 7, characterized in that, The control vector for the m-th stimulus point within time window k is: ； The expression for the stimulus waveform is as follows: ， Or: ， In the formula, p m,k () is the pulse gating function, which takes the value 1 when t falls within the pulse opening interval and takes the value 0 when t falls within the pulse closing interval. (mod) The modulation depth.

9. The transcutaneous electrical stimulation method for autonomic balance regulation according to claim 7, characterized in that, Based on the control vector of the current time window, calculate the comprehensive safety index, including: Based on the stimulus amplitude A in the control vector of the current time window m,k and pulse width τ m,k and the effective electrode contact area A e Calculate the charge density Q using the following formula. mk : ； Based on the control vector of the current time window Charge density Q m,k and current skin temperature (Temp) k The comprehensive safety index C is calculated using the following formula. safe,k : ， In the formula, The maximum permissible stimulation amplitude, maximum permissible pulse width, maximum permissible charge density, and maximum permissible skin temperature are defined as follows:

10. The transcutaneous electrical stimulation method for autonomic balance regulation according to claim 2, characterized in that, The stimulation method further includes an abnormal stopping step during the stimulation process, the abnormal stopping step including: Obtain the pain score for the current time window. ; Based on the pain score within the current time window Heart rate (HR) k and skin conductance level EDA k Calculate the abnormal stop value E according to the following formula. stop,k : ， In the formula, HR lim EDA lim C lim These are the upper limit of safe heart rate, the threshold for abnormal skin conductance, and the upper limit of permissible pain. This is an indicator function that takes the value 1 if the condition is true and 0 otherwise; ∨ represents the logical "OR" operation; When the abnormal stop value E stop,k When the value equals 1, stimulation stops and a safety warning signal is generated.

Images