Wearable device-based traditional Chinese medicine internal medicine deficiency syndrome conditioning effect evaluation method and system

Abstract

The invention discloses a traditional Chinese medicine internal medicine deficiency syndrome conditioning effect evaluation method and system based on a wearable device, and relates to the technical field of traditional Chinese medicine clinical informationization, and the method comprises the following steps: S1, building an electrode skin complex impedance time-frequency fingerprint, executing full-band sweep frequency perturbation, collecting complex impedance data in a continuous monitoring period, and carrying out the detection of the complex impedance time-frequency fingerprint; extracting resonance peak group and sideband coupling path information, and generating a self-excitation risk spectrum as a response reconstruction reference; and S2, performing anti-fact playback based on the self-excitation risk spectrum, replaying the conductance response in a non-resonance interval, comparing a stripping environment drift signal with contact pressure noise through the risk spectrum, and extracting purified anchor point data as an interface feature calibration reference. Abnormal signal recognition and positioning are achieved through complex impedance time-frequency fingerprints, self-excitation risk spectrums, anti-fact playback and micro-capacitance imaging, interference field regulation and control and self-adaptive impedance control are combined, the accuracy and stability of traditional Chinese medicine deficiency syndrome conditioning evaluation are improved, and the method has the high anti-interference capacity and the intelligent characteristic.

Description

Technical Field

[0001] This invention relates to the field of clinical information technology in traditional Chinese medicine, specifically to a method and system for evaluating the therapeutic effects of TCM internal medicine deficiency syndromes based on wearable devices. Background Technology

[0002] "Evaluation of the Effect of Traditional Chinese Medicine (TCM) Internal Medicine Deficiency Syndrome Treatment Based on Wearable Devices" refers to an intelligent health assessment method that integrates modern wearable physiological monitoring technology with TCM syndrome differentiation theory. This method uses wearable devices (such as smart bracelets, chest patches, skin conductance sensors, and sleep monitoring belts) to collect multidimensional physiological signal data in real time, including indicators such as heart rate variability, skin temperature, pulse waveform, respiratory rate, sleep structure, and activity intensity, to reflect the body's Qi and blood circulation, Yin-Yang balance, and organ function. Subsequently, the system combines this objective data with the syndrome characteristics corresponding to TCM internal medicine deficiency syndromes (such as Qi deficiency, blood deficiency, Yin deficiency, and Yang deficiency) to construct a dynamic discrimination model for deficiency and excess and a treatment effect evaluation system. By performing time-series analysis, trend fitting, and feature deviation calculation on the changes in physiological parameters before and after treatment, the system quantitatively presents the promoting effect of TCM treatment interventions (such as acupuncture, medicinal diet, and massage) on the body's homeostasis recovery, realizing the transformation from subjective syndrome differentiation to objective quantification, thereby establishing a scientific evaluation mechanism for TCM deficiency syndrome treatment.

[0003] The existing technology has the following shortcomings: In existing technologies, wearable device-based skin conductance monitoring typically employs fixed-frequency microcurrent stimulation and continuous sampling to acquire changes in human epidermal conductance, reflecting the level of autonomic nervous activity. However, during dynamic monitoring, due to the capacitive coupling effect between the sensing electrodes and the skin interface, self-resonance is highly likely to occur when the monitoring frequency approaches the system's inherent resonant frequency. At this point, the originally minute current fluctuations are amplified in the electrode circuit, forming high-amplitude spurious waveforms. If the system does not have a frequency band suppression window or adaptive amplitude limiting strategy, the high-frequency components of the monitoring signal will be incorrectly identified as physiological stress responses, causing the data processing module to output abnormal "qi stagnation" characteristic signals in the evaluation of the effects of TCM deficiency syndrome treatment. This type of distortion not only disrupts the time-frequency consistency of the evaluation model but also reverses the trend of treatment efficacy judgment, resulting in a serious risk of misjudgment of efficacy.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for evaluating the therapeutic effects of TCM internal medicine deficiency syndromes based on wearable devices, so as to solve the problems in the background art mentioned above.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for evaluating the therapeutic effect of TCM internal medicine deficiency syndrome based on wearable devices, comprising the following steps: S1. Establish the time-frequency fingerprint of electrode skin complex impedance, perform full-band frequency sweep perturbation, collect complex impedance data within the continuous monitoring period, extract the information of resonance peak group and sideband coupling path, and generate self-excited risk spectrum as the response reconstruction benchmark. S2, based on the self-excited risk spectrum, performs counterfactual playback to replay the conductivity response in the non-resonance region. By comparing the risk spectrum, environmental drift signals and contact pressure noise are separated, and purification anchor point data is extracted as the interface feature calibration benchmark. S3, under the constraint of the purification anchor point, perform interface microcapacitance imaging to measure the capacitance gradient change of the skin electrode contact interface, extract the phase lag trajectory and combine it with the purification anchor point data to determine the self-excitation trigger time window. S4, inject a dual-phase conjugate microcurrent mirror sequence around the self-excited triggering time window, superimpose a frequency misalignment micro-perturbation signal, and construct a closed-loop interference field in the time domain to control the energy distribution at the electrode interface; S5 enables the non-Foster adaptive impedance network and inverse diffusion gating mechanism under interference field output conditions, and adjusts the excitation frequency, sampling rhythm and current amplitude upper limit according to the residual density distribution to perform online control and spectrum balancing of the self-excited link.

[0007] Preferably, step S1 includes: A high-stability dry electrode array is installed in the target area to ensure that the electrode-skin contact interface reaches a stable pressing state. The frequency-tunable electrical stimulation unit is connected and full-band frequency sweep perturbation is performed to collect the complex impedance modulus, phase angle, and real and imaginary parts of the complex impedance. The collected complex impedance data are arranged in frequency order, the amplitude change gradient and phase rotation trend are calculated, a complex impedance response surface plot is constructed, the resonance enhancement region is identified, and the coupling path features are extracted. Repeatedly acquire and superimpose multi-cycle complex impedance data under static or low-amplitude dynamic conditions, identify recurring resonance peak groups, merge frequency offset points, and generate complex impedance resonance templates. A self-excited risk spectrum is constructed based on the complex impedance resonant template. The resonant peak group is mapped to a risk level distribution map, and the frequency density, phase change, energy concentration and reproducibility score are output to form a multi-dimensional spectrum structure map.

[0008] Preferably, when constructing the self-excited risk spectrum, frequency density mapping is performed on the resonant peak group identified in the complex impedance resonant template over multiple consecutive monitoring periods, and the frequency bands are classified into high-risk, medium-risk, and low-risk intervals based on the phase change amplitude, energy concentration, and repetition frequency.

[0009] Preferably, step S2 includes: Based on the time-frequency fingerprint of complex impedance and the self-excited risk spectrum, frequency points in the low-risk frequency band within the continuous monitoring period are screened, a multi-frequency resampling sequence is constructed and historical complex impedance data is collected to form a conductivity snapshot matrix. The conductivity snapshot matrix is ​​compared with the complex impedance data collected in the current cycle to extract amplitude difference and phase delay information, generate a difference mapping map, and identify environmental drift and contact pressure interference. Based on the low-offset and highly consistent data points in the difference mapping diagram, phase stability screening and complex impedance stability detection are performed to extract purification anchor points that meet the stability conditions. The purified anchor point set is restructured in the frequency domain to remove data points that overlap with the boundaries of high-risk frequency bands, and an anchor point set with time-frequency uniformity is constructed as a subsequent calibration benchmark.

[0010] Preferably, when constructing the cleanup anchor point set, the anchor points are divided into core anchor point segments and auxiliary anchor point segments based on the distribution density in the frequency dimension. The boundary region is then expanded by interpolation based on the core anchor point segments. At the same time, anchor points that overlap with the boundary positions of high-risk frequency bands are removed to ensure the continuity and stability of the cleanup anchor point set in the frequency domain.

[0011] Preferably, step S3 includes: Based on the low-risk frequency band where the purification anchor point is located, a jump voltage waveform is injected into the electrode array, interface capacitance response data at high time resolution is collected, and frequency tuning is performed simultaneously. Normalization is performed on the collected capacitance data to extract the capacitance gradient change trajectory and identify the abrupt inflection point sequence that falls within the purification anchor point time window. Synchronously collect complex impedance phase response data within the abrupt change inflection point interval, construct phase lag trajectories, and screen out response segments with high overlap and strong repeatability. Based on the time alignment relationship between the phase lag trajectory and the purification anchor point, regions where the lag amplitude and duration meet the standard are identified, forming a self-excitation triggering time window.

[0012] Preferably, step S4 includes: Within the self-excited triggering time window, a mirror excitation sequence is constructed, and positive and negative symmetrical microcurrent pulses are injected to form a conjugate energy structure with complementary amplitudes. By inserting frequency-shifted perturbation signals between mirror sequences, staggered time-domain interference bands are constructed to disrupt the local current concentration trend. By combining mirror sequences with frequency misalignment perturbation signals, a closed-loop interference band structure is constructed, and a periodic time-inversion spectrum is established for continuous energy perturbation modulation. Based on the output results of the interference structure, the transient energy density at the electrode interface is evaluated in real time. High fluctuation nodes are extracted to construct a dynamic reversible time grid. Frequency, amplitude and phase inverse switching operations are performed to achieve energy self-balancing control.

[0013] Preferably, the construction of the dynamic reversible time grid is based on the interference node interval in the closed-loop interference band where the energy density exceeds the warning threshold. Within this interval, the frequency, amplitude and phase of the microcurrent excitation are reversed to keep the energy distribution in a periodic equilibrium state on the time axis and avoid the high-risk self-excitation effect from spreading to other periods.

[0014] Preferably, step S5 includes: Based on the acquisition of complex impedance, phase angle and capacitance gradient differential data at the center position of the interference beat, an instantaneous energy residual spectrum of the electrode interface is constructed to form a continuous periodic residual density sequence. Based on the energy deviation intensity and distribution range in the residual density sequence, a non-Foster adaptive impedance network is used to adjust the excitation frequency, amplitude upper limit and pulse width to achieve negative resistance compensation of the electrode input signal. In the non-Foster regulation process, a time-gated interval is constructed, current amplitude rate and voltage edge slope are limited, and inverse diffusion gating is performed to constrain the energy propagation range. After multi-cycle adjustment, the energy distribution is reconstructed, and the residual change rate and phase trajectory are used to determine whether the system has entered the self-excited link extinction state, thus completing the excitation parameter locking and steady-state observation switching.

[0015] A wearable device-based TCM internal medicine deficiency syndrome treatment efficacy evaluation system includes a complex impedance characteristic modeling module, a counterfactual calibration module, an interface imaging recognition module, an interferometric field modulation module, and an adaptive balance control module. The complex impedance feature modeling module establishes the time-frequency fingerprint of the complex impedance of the electrode skin, performs full-band frequency sweep perturbation, collects complex impedance data within a continuous monitoring period, extracts information on the resonance peak group and sideband coupling path, and generates a self-excited risk spectrum as a response reconstruction benchmark. The counterfactual calibration module performs counterfactual playback based on the self-excited risk spectrum, replays the conductivity response in the non-resonance region, and extracts the purification anchor point data as the interface feature calibration benchmark by comparing the risk spectrum to remove the environmental drift signal and contact pressure noise. The interface imaging and recognition module performs interface microcapacitance imaging under the constraint of the purification anchor point, measures the capacitance gradient change of the skin electrode contact interface, extracts the phase lag trajectory, and determines the self-excitation triggering time window by combining the purification anchor point data. The interference field control module injects a dual-phase conjugate microcurrent mirror sequence around the self-excited triggering time window, superimposes a frequency misalignment micro-perturbation signal, and constructs a closed-loop interference field in the time domain to control the energy distribution at the electrode interface. The adaptive balance control module, under the condition of interference field output, enables the non-Foster adaptive impedance network and the inverse diffusion gating mechanism to adjust the excitation frequency, sampling rhythm and current amplitude upper limit according to the residual density distribution, and performs online control and spectrum balancing of the self-excited link.

[0016] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention achieves early identification of electrode interface resonance risks by constructing a complex impedance time-frequency fingerprint and a self-excitation risk spectrum; it removes environmental disturbances and unstructured contact noise by utilizing counterfactual playback and purification anchor point extraction, providing a calibration benchmark for subsequent signal reconstruction; it achieves precise location of potential abnormal response periods by combining microcapacitance imaging and phase trajectory recognition; furthermore, it constructs a self-suppressive time-domain energy regulation mechanism by forming a closed-loop interference structure through a combination of dual-phase conjugate microcurrents and frequency misalignment disturbances; finally, it achieves active extinguishing of self-excited links and global energy balance of the system during dynamic monitoring through a non-Foster adaptive impedance network and an inverse diffusion gating mechanism. Compared with existing technologies, this method not only improves the time-frequency consistency and quantitative accuracy of the evaluation of the intervention effect of deficiency syndrome treatment, but also enhances the robustness and anti-interference ability of the entire evaluation process in complex scenarios, providing a scientifically feasible technical path for the intelligent and quantitative evaluation of the effect of traditional Chinese medicine treatment. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0018] Figure 1 This is a flowchart of the method for evaluating the therapeutic effect of TCM internal medicine deficiency syndrome based on wearable devices according to the present invention.

[0019] Figure 2 This is a schematic diagram of the module of the TCM internal medicine deficiency syndrome treatment effect evaluation system based on wearable devices according to the present invention. Detailed Implementation

[0020] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0021] This invention provides, for example Figure 1 The method for evaluating the efficacy of TCM internal medicine deficiency syndrome treatment based on wearable devices, as shown, includes the following steps: S1. Establish the time-frequency fingerprint of electrode skin complex impedance, perform full-band frequency sweep perturbation, collect complex impedance data within the continuous monitoring period, extract the information of resonance peak group and sideband coupling path, and generate self-excited risk spectrum for the establishment of benchmark for subsequent response reconstruction. To accurately characterize the dynamic properties of the complex impedance between the skin and electrodes and provide a stable reference basis for subsequent intervention judgments, a complete time-frequency fingerprint generation process for electrode-skin complex impedance needs to be established first. This process systematically constructs a self-excitation risk spectrum through refined electrical stimulation and response acquisition, coupled feature extraction, and spectral modeling, which supports subsequent response reconstruction and judgment criteria formulation. The specific implementation steps are as follows: A highly stable dry electrode array is installed in the target area of ​​the test subject, ensuring a stable pressure connection between each electrode and the skin interface to avoid sudden changes in interface capacitance caused by poor contact or dynamic slippage. The electrodes are made of a bidirectional non-polarized material, featuring low noise and high conductivity, and are connected to a frequency-tunable electrostimulation unit. This unit applies a low-amplitude current signal at the microampere level to the skin surface within a preset safety range, and continuously sweeps the frequency stimulation in a full-band increasing pattern from 1 Hz to 10 kHz. During the frequency sweep, the electrostimulation frequency increases linearly in steps, with each step controlled within 10 Hz to ensure that all changes in the system response are captured in the fine-grained frequency domain. Simultaneously, the response acquisition unit records the complex impedance response at each frequency point, including the impedance magnitude, phase angle, and the real and imaginary parts of the complex impedance. All response signals are quantized by a 24-bit high-resolution analog-to-digital converter and stored in real-time in a data buffer queue according to timestamps. To suppress the effects of environmental power frequency interference and physiological jitter, five rounds of repeated data acquisition were performed at each frequency point, and the five sets of data were weighted and the median was processed to remove possible discrete outliers, thereby enhancing the stability and reproducibility of the frequency domain data.

[0022] After obtaining the complex impedance response sequence in the complete frequency domain, the system proceeds to the resonance feature extraction stage. The system arranges each set of acquired complex impedance data according to its frequency order and calculates the amplitude gradient and phase rotation trend between adjacent frequency bands to identify nonlinear amplification phenomena in the frequency domain response. Specifically, if a significant dip in the real part or a sudden rise in the imaginary part of the complex impedance is observed near the system's natural resonant frequency, it is preliminarily determined that there is a resonance enhancement trend. Based on this, a three-dimensional complex impedance response surface plot is further constructed, with frequency as the horizontal axis, time as the vertical axis, and the complex impedance magnitude as the height, forming a complete time-frequency domain mapping spectrum. In the spectrum, local peak regions represent anomalous amplification points in the complex impedance response. By identifying the frequency bands where these peaks are located and their sideband propagation trajectories, the associated coupling paths can be further deduced, including the capacitive coupling effect at the skin electrode interface and the asymmetric impedance change trend of the deep tissue resistance channel. All high-amplitude response points are marked as preliminary self-excitation risk points, forming an initial frequency cluster set, providing a feature basis for subsequent modeling.

[0023] After initially identifying the suspected self-excited frequency band and its sideband coupling path, data overlay and time-domain stability analysis were performed for continuous monitoring cycles. To this end, the electrode-skin interface was kept in a static or low-amplitude dynamic state, and the frequency sweep stimulation process was repeatedly executed. Complex impedance data were collected at different time points and overlaid for comparison to determine whether the resonant response was a stable, systematic characteristic rather than an occasional anomaly caused by instantaneous fluctuations. In multiple acquisition cycles, high-amplitude response points that repeatedly appeared in the same frequency band were identified as stable self-excited risk peaks, and their frequency distribution center, energy concentration degree, and time drift range were calculated. Based on this, a frequency offset weighting mechanism was introduced to fuse neighboring frequency points with slight frequency shifts to form a more stable resonant peak group characteristic. Finally, a set of complex impedance resonance templates composed of center frequency, phase characteristics, coupling path, and energy threshold was constructed to reflect the resonance tendency and risk region distribution of the skin electrode interface under specific stimulation conditions. This template was combined with the time series to form a cross-cycle time-frequency response set, providing data support for subsequent risk benchmark establishment.

[0024] A self-excited risk spectrum is constructed based on a complex impedance resonant template. During the construction process, the system maps the resonant peak groups identified under different periods onto a risk level distribution map according to their frequency density. Based on their corresponding phase change amplitude, response energy concentration, and reproducibility score, they are divided into high-risk, medium-risk, and low-risk intervals. The high-risk frequency band refers to those that appear stably across multiple periods, with a large phase change rate and extremely low real modulus, easily leading to spurious signal amplification. The medium-risk frequency band refers to those that appear in some periods, exhibiting local amplification and slight shift over time. The low-risk frequency band is the region with a stable response and no obvious amplification trend observed. All risk frequency bands and their characteristic parameters are ultimately encoded into a multi-dimensional spectral structure map, i.e., the self-excited risk spectrum. This risk spectrum serves as a benchmark for complex impedance response reconstruction and can be used for identification, comparison, and screening of actual monitoring data in subsequent stages, and as the primary basis for determining the safe operating area of ​​the electrode interface.

[0025] S2, based on the self-excited risk spectrum, performs counterfactual playback to replay the real electrical conductivity response in the non-resonance region. By comparing the risk spectrum, environmental drift signals and contact pressure noise are separated, and purification anchor point data with time-frequency consistency is extracted for calibration benchmarks for interface feature reconstruction. After completing the complex impedance characteristic modeling and successfully generating the self-excited risk spectrum, in order to eliminate unstructured interference during the monitoring process and provide reliable reference data for the subsequent steady-state reconstruction of interface characteristics, it is necessary to perform a counterfactual playback operation on the historical conductance response during the acquisition process. This operation aims to reconstruct the true conductance behavior under non-resonant conditions and, by comparing it with the risk spectrum, identify reference point data with high time-frequency consistency. Finally, the cleanup anchor point is extracted as the structural calibration input for subsequent phase trajectory identification and trigger window determination. The specific implementation steps are as follows: Based on the established complex impedance time-frequency fingerprint and self-excitation risk spectrum, frequency bands that stably reside in the low-risk range over multiple consecutive monitoring periods were selected as the target range for counterfactual playback. Within these low-risk frequency bands, because they do not exhibit high-risk characteristics such as resonant peak amplification, phase rotation abrupt changes, or frequency clustering, they represent the true conductance response characteristics of the system under non-self-excitation conditions. During the selection process, to avoid misjudgment caused by boundary ambiguity, further boundary narrowing was performed based on the risk change gradient upstream and downstream of the frequency. The spectral center of each low-risk frequency band was used as the selected frequency point, and a multi-frequency resampling sequence was constructed within its ±30 Hz range. Data on the complex impedance magnitude, real part, imaginary part, and phase angle at the same time position within the cycle were collected, forming a multi-dimensional historical conductance snapshot matrix. This snapshot matrix preserves the original conductance response state before self-excitation interference, serving as the basic material for subsequently reconstructing the physiological signal fluctuation trend under "ideal conditions."

[0026] After obtaining the complete historical conductance snapshot matrix, the counterfactual playback operation phase begins. Specifically, the snapshot matrix is ​​compared chronologically with the full-band complex impedance data collected during the current monitoring period to construct a set of time-synchronized reference channels. By comparing the amplitude differences, phase delays, and gradient deviations between the playback values ​​and real-time response values ​​in the low-risk frequency band, a difference map is extracted point-by-point. This map reveals the conductance response trend that the system should exhibit under non-resonant conditions. The deviation from the actual measurement data mainly originates from external disturbances, including skin moisture evaporation, contact pressure changes caused by electrode slippage, and interface capacitance drift due to adjustments in wearing tightness. By comparing the difference map point-by-point with the risk level curve in the risk spectrum, it is possible to effectively identify which conductance response values ​​exhibit significant periodic repetition or frequency band coupling. These deviations are judged as environmental drift or abnormal interface contact signals and are removed from the original data to eliminate their interference with subsequent feature identification results.

[0027] After successfully separating unstructured drift interference from the historical snapshot matrix and real-time data, the cleaned anchor point extraction stage begins. This stage uses identified data points with "low risk, low offset, and high consistency" as the anchoring basis to reconstruct a three-dimensional anchor point grid centered on frequency, time, and phase. During this process, the system performs frequency-band phase stability screening on the cleaned conductivity data, eliminating points where the phase value fluctuation exceeds 1.5 degrees in adjacent sampling periods. Simultaneously, it detects the relative stability rate of change of the real and imaginary parts of the impedance in the time domain, ensuring that the point maintains strong stability and reproducibility not only in the frequency dimension but also in the time dimension. All sampling points that meet the stability conditions are marked as cleaned anchor points and arranged in chronological order, forming a high-confidence, high-fidelity set of cleaned anchor points, providing stable input for interface feature modeling.

[0028] After obtaining the cleaned anchor point set, its distribution in the frequency domain needs to be restructured for subsequent reconstruction of interface response features. In this step, the distribution density of anchor points in the frequency dimension is first statistically analyzed. Based on the density, the anchor points are divided into core anchor point segments and auxiliary anchor point segments. Then, the boundary segments are interpolated and extended based on the core segments to fill the gaps caused by discrete sampling. Next, all anchor points are compared with the center positions of low-risk frequency bands in the previous stage's risk spectrum according to their corresponding time points. Anchor points overlapping with the risk band boundaries are removed to ensure that the anchor points are not affected by potential fluctuations in high-risk frequency bands. The final cleaned anchor point set, with its time-frequency uniformity, phase stability, and interference removal characteristics, can serve as a calibration benchmark in subsequent interface capacitance imaging and phase trajectory reconstruction processes, supporting the determination of dynamic triggering conditions and the identification of energy intervention paths.

[0029] S3 performs interface microcapacitance imaging under the constraint of purification anchor point, measures the capacitance gradient change of the skin electrode contact interface in real time, extracts the phase lag trajectory and combines it with purification anchor point data to determine the potential self-excitation trigger time window, and provides dynamic triggering conditions for the energy intervention stage. After extracting the purification anchor points, to achieve precise identification of the dynamic characteristics of the skin electrode interface and provide accurate temporal positioning for subsequent energy interventions, it is necessary to perform interface microcapacitance imaging based on the purification anchor points. This quantifies the changes in the capacitance gradient at the contact interface and reconstructs the phase hysteresis trajectory, thereby determining the potential self-excitation triggering time window. The specific implementation steps are as follows: With the cleanroom anchor point set established, a micro-capacitance sampling device is activated to perform continuous capacitance measurements at the electrode contact interface with high time resolution. The micro-capacitance sampling device consists of a multi-array conductive film with skin-adhesive elasticity and a programmable excitation source. This excitation source injects short-period, low-amplitude, low-frequency alternating voltage waveforms into the electrode array within the low-risk frequency band corresponding to the cleanroom anchor points, exciting the transient response behavior of the interface capacitance. After each excitation trigger, the measuring end records the dielectric charging and discharging current between the electrode and the skin, and calculates the capacitance value through integration. The sampling frequency is maintained at over 5000 times per second to meet the time resolution requirements for capacitance gradient changes. To ensure data stability and accuracy, the excitation source frequency is tuned in real-time before each measurement, referencing the cleanroom anchor point data, to avoid any known high-risk frequency bands, thereby eliminating interference from resonant responses on the capacitance measurement results. The entire process is executed on a time axis synchronized with the cleanroom anchor points, ensuring temporal consistency of the data and laying the signal foundation for subsequent trajectory extraction.

[0030] After obtaining continuous capacitance measurement data, time series analysis was performed to extract the evolution trend of capacitance gradient changes. To enhance trajectory stability, the raw capacitance data was first baseline normalized, and a sliding window mechanism was used to calculate its local average capacitance value and standard deviation, converting short-term capacitance increments into gradient trajectories. Subsequently, the rate of change of capacitance rise and fall segments within each excitation cycle was tracked along the time axis to identify the inflection points of capacitance gradient changes, i.e., the start and end points of capacitance abrupt changes. These inflection points typically reflect instantaneous adjustments in the interface contact state, which may be caused by minor changes in body position, sweat penetration, temperature drift, or loosening of electrode crimping. Based on the inflection point set, the time interval between adjacent inflection points, the peak capacitance difference, and their corresponding direction of change were further calculated to construct a capacitance gradient change trajectory map. In the trajectory map, special attention was paid to inflection point pairs whose change amplitude exceeded the normal fluctuation threshold and whose time interval fell within the purification anchor sampling window. These trajectories were considered potential causes of phase lag, providing a candidate basis for the next step of lag trajectory extraction.

[0031] After identifying candidate capacitance gradient change trajectories, phase hysteresis trajectories are extracted and quantified. This process uses the capacitance change trajectory as a time-domain reference, while simultaneously acquiring phase response data in conjunction with the stable frequency band marked by the purification anchor point. Within the time window between each pair of capacitance gradient inflection points, the electrode measurement unit is reactivated to sample the phase angle of the complex impedance at high frequencies, recording the phase value every 5 milliseconds and spatially mapping it synchronously with the capacitance gradient trajectory. By comparing the rate of change of the phase angle at different time points, the hysteresis segment of the phase response is identified, i.e., the waveform segment where the phase response begins to deviate from its normal trajectory and then recovers. This type of hysteresis response usually corresponds to the abrupt behavior of the interface capacitance. Furthermore, within the identified phase hysteresis segments, the start time, duration, and maximum hysteresis angle are measured and matched with the position of the purification anchor point on the time axis. Trajectories with high overlap and repeatability are retained, while individual response points with strong drift and sporadic occurrences are eliminated to construct a high-confidence set of phase hysteresis trajectories.

[0032] After obtaining stable and reliable phase hysteresis trajectories, potential self-excitation triggering time windows are determined based on the time matching relationship between this trajectory set and the purification anchor points. Specifically, the starting point of the hysteresis trajectory is aligned with the time nodes of the purification anchor point set to identify time regions where hysteresis responses occur relatively frequently. These regions are then classified into risk levels based on their frequency, hysteresis amplitude, and duration. During this classification, periods with hysteresis amplitudes exceeding 10 degrees, durations exceeding 200 milliseconds, and recurring across multiple monitoring cycles are defined as high-risk self-excitation triggering time windows; while periods with smaller hysteresis amplitudes or fewer occurrences are defined as medium- or low-risk triggering periods. Finally, a complete dynamic triggering interval is formed by extending 50 milliseconds forward and backward from the high-risk triggering time window. This dynamic triggering interval serves as a reference point for initiating energy intervention and can be used to control the timing of subsequent microcurrent injection and feedback path construction, providing a key triggering basis for achieving closed-loop self-suppression intervention.

[0033] S4, inject a dual-phase conjugate microcurrent mirror sequence around the self-excited triggering time window, superimpose the frequency misinterpretation signal, form a closed-loop interference field with self-suppression characteristics in the time domain, construct a reversible time grid to weaken the local transient amplification effect, and realize the stabilization of the energy distribution at the electrode interface. After identifying and determining the self-triggered time window, to prevent abnormal feedback effects caused by local energy amplification at the interface during this period, a dual-phase conjugate microcurrent mirror sequence needs to be actively injected, supplemented by a frequency-shifting perturbation signal to form an interference effect. This constructs a self-suppressing closed-loop interference field on the time axis to stabilize the energy distribution at the electrode-skin contact interface and provide a stable boundary for subsequent spectrum control. The specific implementation steps are as follows: Based on the obtained self-triggered time window, a stimulation timing scheme strictly aligned with this time segment is constructed. Specifically, the window is divided into several micro-stimulation cycles, each no longer than 20 milliseconds, according to its start time, duration, and risk level. Within each cycle, two sets of conjugate microcurrent excitation sequences are set, with amplitudes in the microampere range and pulse widths of 1 to 5 milliseconds, injected into the electrode array in symmetrical forward and reverse waveforms, respectively. These two sets of mirrored microcurrent sequences are precisely arranged in a mirror manner on the time axis, ensuring that the peak of the first sequence aligns with the trough of the second sequence at the same time point, thus forming an amplitude complementarity relationship in the time domain. This mirrored microcurrent input structure does not generate additional load electric field accumulation but instead constitutes a dynamically balanced energy disturbance, suppressing the local potential rise caused by unidirectional charge accumulation and avoiding excessive current concentration at the electrode edge region in the spatial structure, ensuring uniform energy distribution.

[0034] Based on a dual-phase conjugate microcurrent mirror sequence, a frequency-shifting perturbation signal is introduced to enhance the time-domain interference effect. The frequency-shifting perturbation signal refers to a small-amplitude current disturbance signal with a slightly offset frequency but consistent amplitude inserted between the mirror sequences. The frequency offset is controlled between 10 and 30 Hz, and the interference frequency maintains a non-integer multiple relationship with the mirror's main frequency to avoid undesirable resonance caused by periodic superposition. During injection, each set of frequency-shifting perturbation signals is embedded between the original mirror sequences in an independent periodic manner, maintaining symmetry in current amplitude, forming a "phase jitter" structure similar to that between the main interference packet and the secondary perturbation. Through alternating superposition in the time domain, this structure forms several high and low interference bands in a local region. The interference intensity is determined by the phase coincidence and frequency offset, ultimately forming a uniformly dense interference band pattern superimposed on the main axis of the mirror sequence. This type of frequency-shifting perturbation causes a slight shift in the microcurrent excitation at different spatial locations, breaking up the potential charge concentration trend and slowing down the propagation rate of sudden energy fluctuations at the interface, thereby enhancing the energy regulation capability of the original mirror structure.

[0035] During the injection of the mirror structure and the frequency-misaligned perturbation signal, a set of periodic closed-loop interference bands is formed on the time axis. This interference structure, through the staggered positive and negative current peaks and valleys, keeps the total current response at the electrode interface in a near-zero mean state and guides the local energy focus point to shift periodically. This periodic shift keeps the stress response at the electrode-skin interface in a constantly changing dynamic equilibrium, interrupting the transient self-excitation process caused by frequency lock-in or interface capacitive hysteresis. Furthermore, this periodic interference band structure is encoded into a time-domain reversible interferogram, where each interference beat unit contains a positive conjugate pulse, a negative conjugate pulse, and a set of embedded frequency-misaligned perturbation signals, which together constitute a closed-loop interference unit. The interference units are arranged in a repetitive and symmetrical manner in the entire mirror sequence, forming several mutually inverted interference beat sequences on the time axis, constituting a complete time-inversion spectrum. Through this time-inversion spectrum, not only is the sudden amplification effect suppressed, but a rhythmic benchmark is also provided for subsequent nonlinear energy response modeling.

[0036] Based on the stable operation of the closed-loop interferometric field, the energy distribution at the electrode interface is assessed, and a dynamic reversible time grid is constructed. The assessment process involves real-time acquisition of complex impedance changes, phase response delays, and current velocity fluctuations per unit time at the electrode interface to calculate the transient energy density in each micro-region. These energy density values ​​are mapped to the time axis and correspond one-to-one with the closed-loop interferometric beat structure, identifying interferometric node intervals where energy concentration exceeds the warning threshold but remains within the mirror sequence. Based on these node intervals, a dynamic reversible time grid is constructed, where the frequency, amplitude, and phase of the microcurrent excitation automatically switch in reverse according to the interference law within this interval. This ensures that even in the event of unexpected load response fluctuations within a short period, the grid structure can achieve energy self-balancing through symmetry adjustment. The entire time grid structure compresses high-risk self-excitation regions within stable beats, preventing their expansion to other period segments, thereby achieving a continuous and stable energy state at the electrode-skin contact interface, providing a reliable guarantee for subsequent spectrum modulation and feedback triggering.

[0037] S5, under the condition of stable output of the interference field, enables the non-Foster adaptive impedance network and the reverse diffusion gating mechanism to dynamically adjust the excitation frequency, sampling rhythm and current amplitude limit according to the residual density distribution, realize the online extinction and dynamic balance of the self-excited link, and complete the closed-loop self-healing regulation for evaluating the effect of TCM internal medicine deficiency syndrome. Based on the formation and continuous stable output of the closed-loop interference field, to achieve dynamic regulation of the self-excited interference link and stable closed-loop control of the overall energy system, a non-Foster adaptive impedance network and an inverse diffusion gating mechanism need to be further introduced. Based on the real-time observed residual density distribution, the excitation parameters are dynamically adjusted to gradually extinguish potential self-excited paths and achieve energy balance control throughout the entire process. The specific implementation steps are as follows: During the continuous operation of the interferometric field, a high-precision residual monitoring process for the electrode interface state is initiated. This process uses the center position of the closed-loop interferometric beat as a reference, collecting complex impedance response data, transient phase angle changes, and capacitance gradient micro-components at the electrode interface within each beat unit to construct an instantaneous energy residual spectrum. Specifically, the average complex impedance value of the electrodes recorded during the stable output of the interferometric test in the previous cycle is used as a reference baseline. The real-time impedance data at each time point in the current cycle is compared with the baseline, and the energy deviation per unit area is calculated by combining the current input value and frequency parameters corresponding to that point. All deviation values ​​are arranged in chronological order to form a residual density sequence. In this sequence, if certain regions show a rapid increase in energy density or a discontinuous trend in phase response, it indicates that a potential self-excited loop may be triggered in that region, requiring adjustment through dynamic control. To enhance the reliability of the analysis, the residual density data is not sampled individually but is based on the superimposed average of three consecutive cycles, ensuring the stability and representativeness of the identification results and avoiding misjudgments caused by occasional disturbances.

[0038] After identifying regions with high residual density, a non-Foster adaptive impedance network is activated to actively match the electrode interface characteristics for response segments that may cause local instability. The non-Foster impedance structure comprises an active circuit containing negative resistance elements, its main function being to provide adjustable negative resistance compensation for the impedance path of the electrode input excitation signal. The adjustment logic determines the target control segment based on the deviation intensity and distribution range in the aforementioned residual density sequence, and adjusts the equivalent impedance output parameters of the excitation source in real time, making it exhibit negative resistance characteristics in the high residual region to absorb excess energy or provide cancellation current. In implementation, the output frequency of the excitation signal is slightly shifted according to the energy response center of the target region, with each frequency offset controlled to no more than 5 Hz, while simultaneously converging the upper limit of the excitation current amplitude to prevent high-amplitude fluctuations from causing reverse coupling. Furthermore, after the impedance network completes frequency and amplitude adjustment, the corresponding excitation pulse width will also be extended or compressed according to the regional response hysteresis characteristics to match the local energy dissipation rhythm, achieving true dynamic adaptive matching adjustment.

[0039] In the execution of non-Foster adaptive impedance regulation, to prevent the back-diffusion of energy gradients in adjacent regions caused by energy adjustment, a reverse diffusion gating mechanism is introduced to limit the disordered propagation of energy in space. This mechanism constrains the direction of energy flow by constructing a time-gated interval around the trigger point of the non-Foster compensation behavior. Two main strategies are set within the gating interval: first, a current transition rate limit is set to keep the amplitude of the electrode current within a controllable range per unit time, preventing excessively rapid changes that could lead to energy front leakage; second, the edge slope of the voltage waveform is limited to ensure it meets the following capability of the dielectric polarization response, thereby avoiding dielectric excitation anomalies caused by charge hysteresis. Under this mechanism, energy is confined within the target regulation region for local equalization adjustment, preventing its diffusion into low-risk regions and causing system interference. Simultaneously, combined with the stable beat structure provided by the interferometric field, the excitation rhythm is anchored to the starting point of the interferometric beat during the gating process, achieving time-locked synchronization between excitation control and beat modulation, thereby improving the rhythmic consistency and structural controllability of the overall regulation behavior.

[0040] Under the synergistic effect of the non-Foster adaptive impedance network and the inverse diffusion gating mechanism, a multi-cycle global energy dynamic equilibrium process is formed. To monitor the effectiveness of this process, the energy distribution map of the entire electrode interface is reconstructed after each complete adjustment cycle, and the difference between the residual density spectrum of the current cycle and the initial residual spectrum of the reference cycle is analyzed to extract the changing trend of the mean residual value of each high-risk area. When the energy density change rate of all high residual points in the system remains within 5% for two consecutive cycles and no new phase lag trajectory is observed, the current state is marked as "self-excited link extinction state". At this time, all the aforementioned dynamic adjustment parameters are automatically locked at the current value, and the excitation frequency is kept stable through frequency locking operation, transitioning to the steady-state observation stage. In the steady-state stage, the amplitude and sampling rhythm of each excitation cycle are finely adjusted and maintained according to the energy change of the previous cycle to ensure that the system maintains a dynamic equilibrium state in the global dimension for a long time, thereby realizing the closed-loop self-healing regulation capability of the evaluation results in the process of TCM internal medicine deficiency syndrome conditioning, and providing a continuous and stable electrophysiological monitoring basis for the entire conditioning process.

[0041] This invention achieves early identification of electrode interface resonance risks by constructing a complex impedance time-frequency fingerprint and a self-excitation risk spectrum; it removes environmental disturbances and unstructured contact noise by utilizing counterfactual playback and purification anchor point extraction, providing a calibration benchmark for subsequent signal reconstruction; it achieves precise location of potential abnormal response periods by combining microcapacitance imaging and phase trajectory recognition; furthermore, it constructs a self-suppressive time-domain energy regulation mechanism by forming a closed-loop interference structure through a combination of dual-phase conjugate microcurrents and frequency misalignment disturbances; finally, it achieves active extinguishing of self-excited links and global energy balance of the system during dynamic monitoring through a non-Foster adaptive impedance network and an inverse diffusion gating mechanism. Compared with existing technologies, this method not only improves the time-frequency consistency and quantitative accuracy of the evaluation of the intervention effect of deficiency syndrome treatment, but also enhances the robustness and anti-interference ability of the entire evaluation process in complex scenarios, providing a scientifically feasible technical path for the intelligent and quantitative evaluation of the effect of traditional Chinese medicine treatment.

[0042] This invention provides, for example Figure 2 The TCM internal medicine deficiency syndrome treatment effect evaluation system based on wearable devices shown includes a complex impedance characteristic modeling module, a counterfactual calibration module, an interface imaging recognition module, an interferometric field modulation module, and an adaptive balance control module. The complex impedance feature modeling module establishes the time-frequency fingerprint of the complex impedance of the electrode skin, performs full-band frequency sweep perturbation, collects complex impedance data within a continuous monitoring period, extracts information on the resonance peak group and sideband coupling path, and generates a self-excited risk spectrum as a response reconstruction benchmark. The counterfactual calibration module performs counterfactual playback based on the self-excited risk spectrum, replays the conductivity response in the non-resonance region, and extracts the purification anchor point data as the interface feature calibration benchmark by comparing the risk spectrum to remove the environmental drift signal and contact pressure noise. The interface imaging and recognition module performs interface microcapacitance imaging under the constraint of the purification anchor point, measures the capacitance gradient change of the skin electrode contact interface, extracts the phase lag trajectory, and determines the self-excitation triggering time window by combining the purification anchor point data. The interference field control module injects a dual-phase conjugate microcurrent mirror sequence around the self-excited triggering time window, superimposes a frequency misalignment micro-perturbation signal, and constructs a closed-loop interference field in the time domain to control the energy distribution at the electrode interface. The adaptive balance control module, under the condition of interference field output, enables the non-Foster adaptive impedance network and the inverse diffusion gating mechanism to adjust the excitation frequency, sampling rhythm and current amplitude upper limit according to the residual density distribution, and performs online control and spectrum balancing of the self-excited link.

[0043] The method for evaluating the therapeutic effect of TCM internal medicine deficiency syndrome based on wearable devices provided in this embodiment of the invention is implemented through the aforementioned TCM internal medicine deficiency syndrome therapeutic effect evaluation system based on wearable devices. For details of the specific methods and procedures of the TCM internal medicine deficiency syndrome therapeutic effect evaluation system based on wearable devices, please refer to the embodiments of the above-mentioned method for evaluating the therapeutic effect of TCM internal medicine deficiency syndrome based on wearable devices, which will not be repeated here.

[0044] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for evaluating the therapeutic effect of TCM internal medicine deficiency syndromes based on wearable devices, characterized in that, Includes the following steps: S1. Establish the time-frequency fingerprint of electrode skin complex impedance, perform full-band frequency sweep perturbation, collect complex impedance data within the continuous monitoring period, extract the information of resonance peak group and sideband coupling path, and generate self-excited risk spectrum as the response reconstruction benchmark. S2, based on the self-excited risk spectrum, performs counterfactual playback to replay the conductivity response in the non-resonance region. By comparing the risk spectrum, environmental drift signals and contact pressure noise are separated, and purification anchor point data is extracted as the interface feature calibration benchmark. S3, under the constraint of the purification anchor point, perform interface microcapacitance imaging to measure the capacitance gradient change of the skin electrode contact interface, extract the phase hysteresis trajectory and combine it with the purification anchor point data to determine the self-excitation triggering time window. S4, inject a dual-phase conjugate microcurrent mirror sequence around the self-excited triggering time window, superimpose a frequency misalignment micro-perturbation signal, and construct a closed-loop interference field in the time domain to control the energy distribution at the electrode interface; S5 enables the non-Foster adaptive impedance network and inverse diffusion gating mechanism under interference field output conditions, and adjusts the excitation frequency, sampling rhythm and current amplitude upper limit according to the residual density distribution to perform online control and spectrum balancing of the self-excited link.

2. The method for evaluating the therapeutic effect of TCM internal medicine deficiency syndrome based on wearable devices according to claim 1, characterized in that, Step S1 includes: A high-stability dry electrode array is installed in the target area to ensure that the electrode-skin contact interface reaches a stable pressing state. The frequency-adjustable electrical stimulation unit is connected and full-band frequency sweep perturbation is performed to collect the complex impedance modulus, phase angle, and real and imaginary parts of the complex impedance. The collected complex impedance data are arranged in frequency order, the amplitude change gradient and phase rotation trend are calculated, a complex impedance response surface plot is constructed, the resonance enhancement region is identified, and the coupling path features are extracted. Repeatedly acquire and superimpose multi-cycle complex impedance data under static or low-amplitude dynamic conditions, identify recurring resonance peak groups, merge frequency offset points, and generate complex impedance resonance templates. A self-excited risk spectrum is constructed based on the complex impedance resonant template. The resonant peak group is mapped to a risk level distribution map, and the frequency density, phase change, energy concentration and reproducibility score are output to form a multi-dimensional spectrum structure map.

3. The method for evaluating the therapeutic effect of TCM internal medicine deficiency syndrome based on wearable devices according to claim 2, characterized in that, When constructing the self-excited risk spectrum, frequency density mapping is performed on the resonant peak group identified in the complex impedance resonant template over multiple consecutive monitoring periods. The frequency bands are then classified into high-risk, medium-risk, and low-risk ranges based on the phase change amplitude, energy concentration, and repetition frequency.

4. The method for evaluating the therapeutic effect of TCM internal medicine deficiency syndrome based on wearable devices according to claim 1, characterized in that, Step S2 includes: Based on the time-frequency fingerprint of complex impedance and the self-excited risk spectrum, frequency points in the low-risk frequency band within the continuous monitoring period are screened, a multi-frequency resampling sequence is constructed and historical complex impedance data is collected to form a conductivity snapshot matrix. The conductivity snapshot matrix is ​​compared with the complex impedance data collected in the current cycle to extract amplitude difference and phase delay information, generate a difference mapping map, and identify environmental drift and contact pressure interference. Based on the low-offset and highly consistent data points in the difference mapping diagram, phase stability screening and complex impedance stability detection are performed to extract purification anchor points that meet the stability conditions. The purified anchor point set is restructured in the frequency domain to remove data points that overlap with the boundaries of high-risk frequency bands, and an anchor point set with time-frequency uniformity is constructed as a subsequent calibration benchmark.

5. The method for evaluating the therapeutic effect of TCM internal medicine deficiency syndrome based on wearable devices according to claim 4, characterized in that, When constructing the cleanup anchor point set, the anchor points are divided into core anchor point segments and auxiliary anchor point segments based on the frequency dimension distribution density. The boundary region is then expanded by interpolation based on the core anchor point segments. At the same time, anchor points that overlap with the boundary positions of high-risk frequency bands are removed to ensure the continuity and stability of the cleanup anchor point set in the frequency domain.

6. The method for evaluating the therapeutic effect of TCM internal medicine deficiency syndrome based on wearable devices according to claim 1, characterized in that, Step S3 includes: Based on the low-risk frequency band where the purification anchor point is located, a jump voltage waveform is injected into the electrode array, interface capacitance response data at high time resolution is collected, and frequency tuning is performed simultaneously. Normalization is performed on the collected capacitance data to extract the capacitance gradient change trajectory and identify the abrupt inflection point sequence that falls within the purification anchor point time window. Synchronously collect complex impedance phase response data within the abrupt change inflection point interval, construct phase lag trajectories, and screen out response segments with high overlap and strong repeatability. Based on the time alignment relationship between the phase lag trajectory and the purification anchor point, regions where the lag amplitude and duration meet the standard are identified, forming a self-excitation triggering time window.

7. The method for evaluating the therapeutic effect of TCM internal medicine deficiency syndrome based on wearable devices according to claim 1, characterized in that, Step S4 includes: Within the self-excited triggering time window, a mirror excitation sequence is constructed, and positive and negative symmetrical microcurrent pulses are injected to form a conjugate energy structure with complementary amplitudes. By inserting frequency-shifted perturbation signals between mirror sequences, staggered time-domain interference bands are constructed to disrupt the local current concentration trend. By combining mirror sequences with frequency misalignment perturbation signals, a closed-loop interference band structure is constructed, and a periodic time-inversion spectrum is established for continuous energy perturbation modulation. Based on the output results of the interference structure, the transient energy density at the electrode interface is evaluated in real time. High fluctuation nodes are extracted to construct a dynamic reversible time grid. Frequency, amplitude and phase inverse switching operations are performed to achieve energy self-balancing control.

8. The method for evaluating the therapeutic effect of TCM internal medicine deficiency syndrome based on wearable devices according to claim 7, characterized in that, The construction of the dynamic reversible time grid is based on the interference node interval in the closed-loop interference band where the energy density exceeds the warning threshold. Within this interval, the frequency, amplitude and phase of the microcurrent excitation are reversed to keep the energy distribution in a periodic equilibrium state on the time axis and avoid the high-risk self-excitation effect from spreading to other periods.

9. The method for evaluating the therapeutic effect of TCM internal medicine deficiency syndrome based on wearable devices according to claim 1, characterized in that, Step S5 includes: Based on the acquisition of complex impedance, phase angle and capacitance gradient differential data at the center position of the interference beat, an instantaneous energy residual spectrum of the electrode interface is constructed to form a continuous periodic residual density sequence. Based on the energy deviation intensity and distribution range in the residual density sequence, a non-Foster adaptive impedance network is used to adjust the excitation frequency, amplitude upper limit and pulse width to achieve negative resistance compensation of the electrode input signal. In the non-Foster regulation process, a time-gated interval is constructed, current amplitude rate and voltage edge slope are limited, and inverse diffusion gating is performed to constrain the energy propagation range. After multi-cycle adjustment, the energy distribution is reconstructed, and the residual change rate and phase trajectory are used to determine whether the system has entered the self-excited link extinction state, thus completing the excitation parameter locking and steady-state observation switching.

10. A system for evaluating the therapeutic effect of TCM internal medicine deficiency syndromes based on wearable devices, used to implement the method for evaluating the therapeutic effect of TCM internal medicine deficiency syndromes based on wearable devices as described in any one of claims 1-9, characterized in that, It includes a complex impedance characteristic modeling module, a counterfactual calibration module, an interface imaging recognition module, an interferometric field manipulation module, and an adaptive balance control module. The complex impedance feature modeling module establishes the time-frequency fingerprint of the complex impedance of the electrode skin, performs full-band frequency sweep perturbation, collects complex impedance data within a continuous monitoring period, extracts information on the resonance peak group and sideband coupling path, and generates a self-excited risk spectrum as a response reconstruction benchmark. The counterfactual calibration module performs counterfactual playback based on the self-excited risk spectrum, replays the conductivity response in the non-resonance region, and extracts the purification anchor point data as the interface feature calibration benchmark by comparing the risk spectrum to remove the environmental drift signal and contact pressure noise. The interface imaging and recognition module performs interface microcapacitance imaging under the constraint of the purification anchor point, measures the capacitance gradient change of the skin electrode contact interface, extracts the phase lag trajectory, and determines the self-excitation triggering time window by combining the purification anchor point data. The interference field control module injects a dual-phase conjugate microcurrent mirror sequence around the self-excited triggering time window, superimposes a frequency misalignment micro-perturbation signal, and constructs a closed-loop interference field in the time domain to control the energy distribution at the electrode interface. The adaptive balance control module, under the condition of interference field output, enables the non-Foster adaptive impedance network and the inverse diffusion gating mechanism to adjust the excitation frequency, sampling rhythm and current amplitude upper limit according to the residual density distribution, and performs online control and spectrum balancing of the self-excited link.

Images