Electric shock simulation training detection and evaluation method, system and applet

Abstract

The invention discloses an electric shock simulation training detection and evaluation method, system and applet, and relates to the technical field of electric shock simulation training, and the method comprises the following steps: selecting an electrical stimulation instrument, applying a preset controllable electrical stimulation signal to a designated muscle area, and simulating an electrical signal of a real external environment under a precise control condition; the electric field change of muscular tissues in a delimited area is directly monitored through a micro field effect sensor, and the diffusion trend of charges is reflected in real time; electric field change data obtained in real time is preprocessed by applying a signal processing technology, then key features reflecting charge accumulation are extracted, and deep analysis is performed on the extracted key features. The electric field change in the muscular tissue is monitored in real time, the charge accumulation risk is recognized in advance through the deep learning model, the pulse duty ratio is automatically adjusted when the risk occurs, abnormal polarization and electrolytic reaction are effectively prevented, and the risks of tissue damage, cell necrosis and nerve damage are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of electric shock simulation training, and in particular to an electric shock simulation training detection and evaluation method, system and applet. Background Art

[0002] Electric shock simulation training detection is an experimental method based on electrical stimulation, which is often used in fields such as neuroscience, psychology, and biomedical engineering to study the response of the nervous system to electrical stimulation, or for neurorehabilitation training. In this method, researchers use controllable electrical stimulation to simulate electrical signals in the external environment to stimulate the nerves, muscles, or specific areas of the brain of the subjects to induce specific physiological or behavioral responses, and record relevant data through biosensors or other detection equipment. This technology can be used to evaluate neural function, detect neural damage, train neural adaptive recovery, or study individual sensitivity to electrical stimulation and stress response.

[0003] The prior art has the following deficiencies: When controllable electrical stimulation is used to simulate external electrical signals and applied to the muscles of the subjects, if the charge cannot diffuse normally and accumulates, resulting in excessive local potential difference, it may cause abnormal polarization effects or electrolytic reactions, causing a series of physiological abnormalities such as cell membrane ion channel disorders and metabolic imbalances. Since existing technologies are still difficult to monitor charge accumulation in real time and make effective adaptive adjustments, once the above situation occurs, it often leads to tissue damage, electrical burns, cell necrosis, and even further develops into long-term damage to the nervous system or chronic pain syndrome, thus posing serious risks to patients' rehabilitation training and daily functional recovery.

[0004] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0005] The purpose of the present invention is to provide a method, system and applet for detecting and evaluating electric shock simulation training, which can effectively prevent abnormal polarization and electrolytic reactions, reduce the risk of tissue damage, cell necrosis and nerve damage by real-time monitoring of electric field changes in muscle tissue, using a deep learning model to identify the risk of charge accumulation in advance, and automatically adjust the pulse duty cycle when the risk occurs. This system based on multi-level data processing and adaptive regulation can not only ensure that the charge is evenly diffused in the muscle, but also dynamically optimize the electrical stimulation parameters during the treatment process, thereby making rehabilitation training and other electrical stimulation treatments more accurate and safe, and helping to improve the long-term rehabilitation effect of patients, so as to solve the problems in the above-mentioned background technology.

[0006] In order to achieve the above object, the present invention provides the following technical solution: a method for detecting and evaluating electric shock simulation training, comprising the following steps: Select an electrical stimulation device and apply a pre-set controllable electrical stimulation signal to the designated muscle area to simulate the electrical signal of the real external environment under precise control conditions; The electric field changes of muscle tissue in the defined area are directly monitored by micro field effect sensors, reflecting the diffusion trend of charge in real time. Signal processing technology is used to pre-process the electric field change data acquired in real time, and then the key features reflecting charge accumulation are extracted. The extracted key features are deeply analyzed to characterize the risk of charge accumulation, and a standardized feature vector is constructed for input into the deep learning model. The constructed feature vector is input into a pre-trained and fully verified deep learning model, the real-time feature vector currently input is analyzed, and the prediction result of potential charge accumulation risk in the charge diffusion trend is output to achieve early identification and early warning of charge accumulation risk; Once the deep learning model issues an accumulation risk warning, it automatically enters the real-time risk assessment mode and immediately uses the data provided by the micro field effect sensor to accurately calculate the real-time change value of the potential difference between different locations within the defined area and quantify the current severity of charge accumulation. After quantitatively calculating the potential difference, the pulse duty cycle adjustment action is automatically performed according to the intelligent control strategy. When the potential difference exceeds the preset potential difference threshold, the "low" state time of the pulse signal is extended, that is, the pulse interval is increased, thereby reducing the total amount of charge injection per unit time and alleviating the degree of local accumulation.

[0007] Preferably, the demarcated area refers to the predetermined range of muscles or tissues that are directly acted upon by electrodes or affected by charge accumulation during a muscle electrical stimulation experiment or treatment, and is precisely selected based on the anatomical structure, electrical stimulation target, clinical treatment needs, or experimental design.

[0008] Preferably, features reflecting the charge accumulation situation are extracted, wherein the extracted features include the cumulative values ​​of the time required for the potential to return to a baseline level and the rate of change of the potential, and an in-depth analysis is performed on the cumulative values ​​of the time required for the potential to return to a baseline level and the rate of change of the potential, and a potential regression delay reference value and a charge accumulation rate reference value are generated respectively, and the charge accumulation risk is characterized by the potential regression delay reference value and the charge accumulation rate reference value, and a standardized feature vector for input into a deep learning model is constructed at the same time.

[0009] Preferably, the constructed potential regression delay reference value and charge accumulation rate reference value are input into a pre-trained and fully verified deep learning model, the real-time feature vector of the current input is analyzed, and the accumulation assessment coefficient is output through the deep learning model. Based on the accumulation assessment coefficient, early identification and warning of potential charge accumulation risks in the charge diffusion trend are carried out.

[0010] Preferably, the accumulation assessment coefficient generated when predicting the potential charge accumulation risk in the charge diffusion trend through the deep learning model is compared and analyzed with the preset accumulation assessment coefficient reference threshold, so as to carry out early identification and early warning of the charge accumulation risk. The specific steps are as follows: If the accumulation assessment coefficient is greater than the accumulation assessment coefficient reference threshold, a risk signal is generated, indicating that there is a risk of charge accumulation in the electric field changes of the muscle tissue in the designated area, and an accumulation risk warning is issued; if the accumulation assessment coefficient is less than or equal to the accumulation assessment coefficient reference threshold, a normal change signal is generated, indicating that the electric field of the muscle tissue in the designated area is changing normally, and no risk warning is issued.

[0011] Preferably, the real-time change value of the potential difference between different positions within the defined area is accurately calculated through the data provided by the micro field effect sensor to quantify the severity of the current charge accumulation. The specific steps are as follows: When the deep learning model issues an accumulation risk warning, it automatically enters the real-time risk assessment mode and uses the micro field effect sensor to measure the potential data at different locations in the designated area. The potential difference in the designated area is calculated by the instantaneous electric field change value collected by the micro field effect sensor, and the accumulation assessment coefficient is combined for quantitative analysis to generate a charge accumulation severity index to determine the severity of the charge accumulation. The calculation formula of the charge accumulation severity index is as follows: , in: is an indicator of the severity of charge accumulation, It is the accumulation assessment coefficient output by the deep learning model, which is used to measure the risk level of charge accumulation. is the reference threshold of the accumulation evaluation coefficient, is the maximum potential difference at different positions within the demarcated area, indicating the degree of local polarization caused by charge accumulation in the demarcated area. The calculation expression is: , in: and are the maximum and minimum electric field strengths in the demarcated area, respectively.

[0012] Preferably, the pulse duty cycle adjustment action is automatically performed according to the intelligent control strategy, and the specific steps are as follows: Based on the calculated charge accumulation severity index , an adaptive pulse duty cycle control strategy is adopted to ensure uniform charge diffusion. The control expression is as follows: , in: is the adjusted pulse duty cycle used to reduce the effect of charge accumulation, is the reference pulse duty cycle, i.e. the default pulse duty cycle under normal circumstances. It is the adjustment coefficient that controls the sensitivity of duty cycle adjustment.

[0013] An electric shock simulation training detection and evaluation system, including an electric stimulation signal application module, an electric field monitoring module, a signal processing and feature extraction module, a deep learning risk prediction module, a real-time risk assessment module, and an adaptive pulse control module; The electrical stimulation signal application module applies a pre-set controllable electrical stimulation signal to the designated muscle area, simulating the electrical signal of the real external environment under precise control conditions; The electric field monitoring module directly monitors the electric field changes of muscle tissue in the defined area through micro field effect sensors, reflecting the diffusion trend of charge in real time; The signal processing and feature extraction module uses signal processing technology to pre-process the electric field change data acquired in real time, then extracts the key features reflecting charge accumulation, conducts in-depth analysis on the extracted key features, characterizes the risk of charge accumulation, and constructs a standardized feature vector for input into the deep learning model; The deep learning risk prediction module inputs the constructed feature vector into a pre-trained and fully verified deep learning model, analyzes the current real-time feature vector input, and outputs the potential charge accumulation risk prediction results in the charge diffusion trend, thus realizing early identification and early warning of charge accumulation risks; The real-time risk assessment module automatically enters the real-time risk assessment mode once the deep learning model issues an accumulation risk warning. It immediately uses the data provided by the micro field effect sensor to accurately calculate the real-time change value of the potential difference between different locations within the defined area and quantify the current severity of charge accumulation. After quantitatively calculating the potential difference, the adaptive pulse control module automatically performs pulse duty cycle adjustment according to the intelligent control strategy. When the potential difference exceeds the preset potential difference threshold, the "low" state time of the pulse signal is extended, that is, the pulse interval is increased, thereby reducing the total amount of charge injection per unit time and alleviating the degree of local accumulation.

[0014] In the above technical solution, the technical effects and advantages provided by the present invention are: By real-time monitoring of the electric field changes in muscle tissue, using deep learning models to identify the risk of charge accumulation in advance, and automatically adjusting the pulse duty cycle when the risk occurs, it effectively prevents abnormal polarization and electrolytic reactions, and reduces the risk of tissue damage, cell necrosis and nerve damage. This system based on multi-level data processing and adaptive regulation can not only ensure that the charge is evenly diffused in the muscle, but also dynamically optimize the electrical stimulation parameters during the treatment process, making rehabilitation training and other electrical stimulation treatments more accurate and safe, and helping to improve the long-term rehabilitation effect of patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0016] Figure 1 The present invention is a method flow chart of an electric shock simulation training detection and evaluation method.

[0017] Figure 2 The present invention is a schematic diagram of a module of an electric shock simulation training detection and evaluation system. DETAILED DESCRIPTION

[0018] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0019] The present invention provides Figure 1 The electric shock simulation training detection and evaluation method shown comprises the following steps: Select an electrical stimulation device to apply a pre-set controllable electrical stimulation signal to the designated muscle area through externally attached or implanted electrodes, simulating the electrical signal of the real external environment under precise control conditions; Stimulation parameters include stimulation voltage, current, pulse frequency, pulse waveform and initial duty cycle. The initial parameter setting is optimized based on the subject's physiological characteristics, treatment objectives and safety requirements. The electrical signals of the real external environment are simulated under precise control conditions, laying the foundation for subsequent charge monitoring, feature extraction and dynamic regulation.

[0020] The electric field changes of muscle tissue in the defined area are directly monitored by micro field effect sensors, reflecting the diffusion trend of charge in real time. The defined area refers to a specific muscle or tissue area that is pre-determined to be directly acted upon by electrodes or that may be affected by charge accumulation during muscle electrical stimulation experiments or treatments. This area is usually precisely selected based on anatomical structure, electrical stimulation target, clinical treatment needs, or experimental design.

[0021] Based on the anatomical structure, it means that it is necessary to refer to the distribution of muscles, nerves, blood vessels and surrounding tissues to select the most suitable location for electrode placement to ensure that the stimulation signal can effectively act on the target tissue while avoiding unnecessary side effects or interference. Based on the target of electrical stimulation, it means selecting the best stimulation area according to the specific application scenario of electrical stimulation (such as muscle rehabilitation, nerve regulation, motor enhancement, etc.). For example, for motor rehabilitation of hemiplegic patients, the quadriceps or upper limb flexor muscles may be selected as the demarcated area to enhance their contraction ability. Based on clinical treatment needs, it means that in medical applications, according to the characteristics of the patient's symptoms, such as muscle atrophy, nerve damage or pain management, the part that can best improve symptoms is selected. For example, electrical stimulation for the treatment of chronic pain may select specific nerve nodes, while stimulation for motor recovery may target specific muscle groups. Based on experimental design, it means that in the process of scientific research or technical optimization, in order to study the law of charge diffusion, electric field distribution characteristics or optimize stimulation parameters, researchers may select specific muscles or nerve areas as research objects, and set up reasonable control groups and variable controls to ensure the comparability and scientificity of experimental data. Only by comprehensively considering these factors can the stimulation area be accurately defined, the electrical stimulation effect be optimized, and adverse reactions be minimized.

[0022] Signal processing technology is used to pre-process the electric field change data acquired in real time, and then the key features reflecting charge accumulation are extracted. The extracted key features are deeply analyzed to characterize the risk of charge accumulation, and a standardized feature vector is constructed for input into the deep learning model. When applying signal processing technology to pre-process the electric field change data acquired in real time, a variety of methods are usually used to ensure data quality and accuracy. These include filtering (such as low-pass filtering, high-pass filtering, and band-pass filtering) to remove high-frequency noise and interference signals, smooth the data curve, and enhance the main features of the signal; in addition, denoising technology (such as wavelet denoising, Kalman filtering) can further reduce the impact of random noise on the data; at the same time, normalization and standardization processing may also be performed so that data collected in different time periods or different sensors have a unified scale, which is convenient for subsequent feature extraction and model training; in addition, time-frequency analysis methods such as Fourier transform and wavelet transform can convert signals from the time domain to the frequency domain, revealing the periodicity and transient characteristics in the data, which helps to accurately capture the dynamic changes of charge accumulation. Through these comprehensive signal processing technologies, the stability and reliability of real-time data can be ensured, providing a solid foundation for the subsequent extraction of key features and deep learning model prediction.

[0023] Features reflecting the charge accumulation situation are extracted, wherein the extracted features include the cumulative value of the time required for the potential to return to the baseline level and the cumulative value of the potential change rate. The time required for the potential to return to the baseline level and the cumulative value of the potential change rate are deeply analyzed to generate a potential regression delay reference value and a charge accumulation rate reference value, respectively. The charge accumulation risk is characterized by the potential regression delay reference value and the charge accumulation rate reference value, and a standardized feature vector is constructed for input into the deep learning model.

[0024] When the time required for the potential to return to the baseline level is longer than the normal recovery time, it usually means that there is a potential risk of charge accumulation in the muscle tissue within the demarcated area. Under normal circumstances, after the end of electrical stimulation, the potential in the tissue should return to a steady state in a short period of time, which mainly depends on tissue conductivity, ion diffusion capacity, membrane potential recovery mechanism, and the charge redistribution capacity of the surrounding environment. However, if the recovery time is significantly prolonged, it may indicate that the charge diffusion in the local area is limited, resulting in the retention of residual potential. This phenomenon may be due to charge shielding effects, abnormal cell membrane polarization, ion channel disorders, or changes in tissue electrical impedance, thereby affecting the normal physiological function of muscle tissue. Long-term charge accumulation may also cause cell membrane potential imbalance, abnormal tissue metabolism, and even lead to chronic inflammation or decreased stimulation tolerance.

[0025] The specific steps for in-depth analysis of the time required for the potential to return to the baseline level and generate a reference value for the potential return delay are as follows: First, a potential regression time curve is established to describe the dynamic changes of the electric field returning to the baseline. The construction expression of the potential regression time curve is as follows: , in: is the potential regression time curve, which represents the nonlinear change of the potential recovery process. represents the potential value measured at time t, is the rate of change of potential, reflecting how fast the potential recovers. and They represent the moment when the electrical stimulation ends and the moment when the potential recovers to the set threshold, respectively, and are used to limit the calculation interval. is a nonlinear adjustment parameter, if , then the effect of the drastic change in the potential recovery rate is enhanced. If , then the small change in the smooth potential recovery rate is is the time attenuation factor, which is used to reduce the impact far from the end point of electrical stimulation and give greater weight to the early recovery stage; This step calculates the weighted cumulative value of the potential recovery rate by integrating it, combined with the exponential decay factor , ensuring that the early potential changes have a more significant impact, while avoiding the interference of small fluctuations at the far end on the calculation results. This process can accurately capture the abnormal recovery characteristics caused by charge shielding, polarization lag, etc. during the potential regression process, and provide a basis for the construction of the subsequent potential regression delay reference value.

[0026] After the potential regression time curve is obtained, the potential regression delay reference value representing the risk of charge accumulation is generated through the potential regression time curve. The generated expression is: , in: is the potential regression delay reference value, is the scale constant of the potential recovery time, normalizing the recovery time range of different individuals. is an exponential scaling parameter used to adjust the influence of the recovery time on the potential regression delay reference value. The baseline electrical potential of muscle tissue within the demarcated area at rest.

[0027] Add the logarithmic normalization term, that is This term is used to avoid excessive amplification of the potential regression delay reference value in extreme cases, while emphasizing the impact of abnormal potential retention. The accumulation of potential deviation is calculated in combination with time integration to ensure that the potential regression delay reference value not only depends on the recovery time, but also takes into account the cumulative effect of potential anomalies during the entire recovery process.

[0028] It can be seen from the potential regression delay reference value that the larger the performance value of the potential regression delay reference value generated by in-depth analysis of the time required for the potential to return to the baseline level, the greater the risk of charge accumulation in the electric field changes of the muscle tissue in the delineated area, and vice versa. The potential regression delay reference value is generated by in-depth analysis of the time required for the potential to return to the baseline level and the cumulative effect of abnormal deviations during the potential recovery process. The larger the performance value, the slower the local potential recovery and the more obvious the abnormal phenomenon, thereby indicating that the risk of charge accumulation in the muscle tissue in the area is greater; conversely, a lower potential regression delay reference value means faster potential recovery and fewer abnormal deviations, indicating a lower risk of charge accumulation.

[0029] The electric field changes in the muscle tissue in the defined area are monitored by a micro field effect sensor. If the cumulative value of the rate of potential change continues to rise, it indicates that the charge in the area has not been effectively diffused and there is a risk of potential charge accumulation. This phenomenon usually means that the local potential continues to increase or maintain a high value in a short period of time, and fails to return to the baseline level according to the normal charge diffusion pattern. Under normal circumstances, the charge should diffuse rapidly in the tissue and tend to stabilize after electrical stimulation. If the rate of potential change continues to accumulate and rise, it may be due to the continuous accumulation of charge in the local area and the failure to effectively dissipate through natural diffusion or neutralization mechanisms. This may lead to abnormal polarization effects, affect the normal function of muscle and nerve cells, and may cause tissue damage, electrolytic reactions, or chronic stimulation adaptation problems.

[0030] The specific steps for in-depth analysis of the accumulated value of the potential change rate to generate a reference value of the charge accumulation rate are as follows: In the muscle tissue within the defined area, the charge accumulation is characterized by the cumulative trend of the rate of change of the electric field. For spatial location In order to quantify the charge accumulation rate, the instantaneous cumulative value of the potential change rate is first calculated. The calculation expression is: , in: is the instantaneous cumulative value of the potential change rate, which characterizes the intensity of the electric field change and its distribution trend in space. To define the area (three-dimensional monitoring area), that is, the electric field monitoring range within the muscle tissue, Along the local potential direction The rate of change of the potential gradient characterizes how the local electric field changes with the stimulation conditions. is the spatial attenuation factor, where represents the distance from the point to the stimulation electrode, To adjust the adjustment coefficient of the spatial weight, ensure that the area far away from the stimulus source does not contribute too much to the overall potential change, Indicated in The tiny volume unit of internal integration; This step calculates the rate of change of electric potential in the monitoring area through integration, and combines it with exponential decay weights to make the area close to the stimulus source contribute more to the overall charge accumulation rate. At the same time, the differences in electric field changes at different locations are taken into account, thereby providing a spatially weighted instantaneous charge change rate representation for subsequent calculations.

[0031] In order to more accurately assess the risk of charge accumulation, the instantaneous accumulation value of the potential change rate Based on this, a nonlinear growth factor is introduced to construct a reference value of charge accumulation rate. The constructed expression is: , in: It is a reference value of charge accumulation rate, which is used to characterize the risk of charge accumulation in a local area. The larger the value, the higher the degree of charge accumulation. To balance the parameters, avoid the denominator approaching zero, and ensure the stability of the exponential calculation, is the divergence of current density, which characterizes the flow of charge in the tissue. If its value is large, it indicates that there is a strong inflow or outflow of charge in the area. The square of the current density divergence in the entire monitoring area is integrated to measure the unevenness of local charge movement; if this value is small, it means that the charge diffusion is more uniform and the accumulation risk is low. A hyperbolic sine function is used to enhance the nonlinear response, where Controlling the effect of the rate of potential change on the final charge accumulation rate reference ensures that the response can be amplified when charge accumulation rapidly intensifies.

[0032] It should be noted that ,in, , , They represent the current density at x , y , the component in the z direction, , , They represent the current density at x , y , the rate of change in the z direction. Current density refers to the amount of current flowing per unit area and is used to describe the distribution of current in space.

[0033] It can be seen from the charge accumulation rate reference value that the larger the performance value of the charge accumulation rate reference value generated by in-depth analysis of the cumulative value of the potential change rate, the greater the risk of charge accumulation in the electric field change of the muscle tissue in the delimited area, and vice versa. The charge accumulation rate reference value directly reflects the accumulation of the potential change rate and is used to measure the charge accumulation risk of local muscle tissue. The larger the performance value, the higher the rate of change of the local electric field, and the charge diffusion may be hindered, resulting in continuous accumulation of charge and failure to diffuse or dissipate in time, which may cause abnormal polarization effects, electrolytic reactions or local tissue damage. On the contrary, if the charge accumulation rate reference value is small, it means that the potential change rate is low, the charge can diffuse evenly and reach a steady state, which means that the risk of charge accumulation is low and the electrical stimulation environment of the muscle tissue is safer.

[0034] Input the constructed feature vector into a pre-trained and fully verified deep learning model (such as LSTM recurrent neural network, Transformer model, CNN or hybrid model), analyze the current real-time feature vector input, and output the potential charge accumulation risk prediction result in the charge diffusion trend to achieve early identification and warning of charge accumulation risk; The constructed potential regression delay reference value and charge accumulation rate reference value are input into a pre-trained and fully verified deep learning model (such as LSTM recurrent neural network, Transformer model, CNN or hybrid model), the real-time feature vector of the current input is analyzed, and the accumulation assessment coefficient is output through the deep learning model. Based on the accumulation assessment coefficient, early identification and warning of potential charge accumulation risks in the charge diffusion trend are carried out.

[0035] A pre-trained and well-validated deep learning model refers to a neural network model that has been fully trained and optimized on a large number of high-quality data sets. The model has stable generalization capabilities, can accurately identify the risk of charge accumulation, and predict the rate of potential change and charge diffusion trends. Since charge diffusion involves complex time series dynamics, spatial distribution characteristics, and nonlinear signal patterns, using deep learning to automatically extract features and make predictions is one of the best options. Deep learning models, such as LSTM (Long Short-Term Memory Network), Transformer, CNN (Convolutional Neural Network), or their hybrid architectures, can learn different charge diffusion patterns from a large amount of electric field change data and predict potential charge accumulation trends based on historical data.

[0036] In order to make the model reliable and accurate, sufficient pre-training and validation are necessary. Pre-training refers to the use of a large amount of labeled data to conduct preliminary training of the neural network so that it can learn basic patterns, such as early signals of charge accumulation, normal diffusion characteristics, and abnormal electric field behavior. Rigorous Validation ensures the generalization ability of the model through a variety of evaluation methods (such as cross-validation, independent test set evaluation, and real experimental comparison), that is, the model can not only perform well on the training data, but also maintain stable prediction capabilities on new data. In this process, a variety of optimization techniques may be used, such as hyperparameter tuning, regularization, data augmentation, and adaptive learning rate optimization to improve the robustness and adaptability of the model. Ensure that the model can not only accurately identify known charge accumulation patterns, but also effectively detect new and unseen anomalies, so that it has the ability to be applied in real scenarios.

[0037] LSTM (Long Short-Term Memory) is suitable for processing time series data. It can memorize long-term charge diffusion patterns and capture accumulation trends caused by delayed potential changes. For example, during muscle electrical stimulation, charge accumulation may not be manifested immediately, but gradually accumulates over time, which requires a recursive neural network (RNN) structure such as LSTM to identify long-term accumulation trends. Through the gating mechanism, LSTM can filter out irrelevant information when analyzing the potential regression delay reference value and the charge accumulation rate reference value, retaining only important features related to charge diffusion, thereby providing more accurate risk prediction.

[0038] Transformer models (such as the Time-Series Transformer for time series prediction) provide an architecture based on the attention mechanism (Self-Attention Mechanism) that can establish long-distance dependencies between different time steps. In charge diffusion modeling, Transformer can effectively process spatial potential change data collected by multiple sensors, identify charge exchange patterns between different muscle regions, and integrate multi-dimensional data to provide a more comprehensive risk assessment. In addition, Transformer can also combine the potential regression delay reference value and the charge accumulation rate reference value to identify charge anomalies that change dramatically in a short period of time and predict whether accumulation will occur in the future.

[0039] CNN (Convolutional Neural Network) is mainly used to extract the spatial characteristics of charge diffusion. In the process of muscle electric field monitoring, CNN can be used to analyze the charge distribution in different regions and detect abnormal charge accumulation points. In particular, when charge accumulation appears as local high potential areas, CNN can identify the spatial patterns of these areas through convolutional layers and combine with other deep learning networks for trend prediction. In addition, CNN can also be combined with LSTM or Transformer to achieve spatio-temporal modeling, and simultaneously analyze the risk of charge accumulation in time and space dimensions.

[0040] In order to ensure the stability and efficiency of the deep learning model, a hybrid model is usually used, combining CNN+LSTM or Transformer+LSTM, using CNN to extract spatial features, LSTM to process time series information, and Transformer to establish long-range dependencies, to obtain the most comprehensive charge diffusion prediction capabilities. For example, an efficient hybrid model architecture may be as follows: Input layer: input potential regression delay reference value, charge accumulation rate reference value; CNN extracts spatial features: The convolutional layer analyzes the distribution of electrical potential at different locations of muscle tissue; LSTM identifies temporal dynamic features: analyzes charge diffusion patterns over a period of time and predicts future trends; Transformer performs global pattern learning: integrating the results of CNN and LSTM, establishing long-term dependencies, and optimizing risk assessment of charge accumulation; Output layer: Generates accumulation evaluation coefficients and provides prediction results of charge diffusion trends.

[0041] Through this deep learning architecture, the system is able to issue real-time warnings when potential charge accumulation risks are detected.

[0042] The deep learning model is not specifically limited here, and can achieve the delay reference value of the potential regression and charge accumulation rate reference value Perform comprehensive analysis to generate accumulation evaluation coefficients In order to realize the technical solution of the present invention, the present invention provides a specific implementation method; the accumulation evaluation coefficient The resulting calculation formula is: , where , are the potential regression delay reference values and charge accumulation rate reference value The preset scaling factor of , All are greater than 0. The preset proportional coefficients here refer to the pre-set weight parameters in order to reasonably balance the influence of the potential regression delay reference value and the charge accumulation rate reference value on the final evaluation result when calculating the accumulation evaluation coefficient. Their role is to adjust the contribution of different features to the final accumulation evaluation coefficient during the calculation process to reflect the relative importance of each indicator to the charge accumulation risk.

[0043] Specifically: like , it means that the reference value of potential regression delay has a greater weight in the accumulation risk assessment, that is, it is more inclined to believe that the delay of potential regression is the main factor of accumulation risk.

[0044] like , it means that the charge accumulation rate reference value has a stronger impact on the accumulation risk, that is, the system pays more attention to the change in the rate of charge accumulation.

[0045] Preset scale factor and satisfy , Both are greater than 0 to ensure that the potential regression delay reference value and the charge accumulation rate reference value contribute to the results in the calculation and are not completely ignored.

[0046] These coefficients can usually be set through experimental data statistics, machine learning optimization or empirical adjustment to ensure that the accumulation assessment coefficient can accurately and reasonably reflect the charge accumulation risk and is applicable to different individuals and application scenarios.

[0047] It can be seen from the accumulation assessment coefficient that the greater the performance value of the potential regression delay reference value generated by in-depth analysis of the time required for the potential to return to the baseline level, the greater the performance value of the charge accumulation rate reference value generated by in-depth analysis of the cumulative value of the potential change rate. In other words, the greater the performance value of the accumulation assessment coefficient generated when predicting the potential charge accumulation risk in the charge diffusion trend through the deep learning model, the greater the risk of charge accumulation in the electric field changes of the muscle tissue in the delineated area, and vice versa.

[0048] The accumulation assessment coefficient generated when predicting the potential charge accumulation risk in the charge diffusion trend through the deep learning model is compared and analyzed with the preset accumulation assessment coefficient reference threshold to perform early identification and early warning of the charge accumulation risk. The specific steps are as follows: If the accumulation assessment coefficient is greater than the accumulation assessment coefficient reference threshold, a risk signal is generated, indicating that there is a risk of charge accumulation in the electric field changes of the muscle tissue in the designated area, and an accumulation risk warning is issued; if the accumulation assessment coefficient is less than or equal to the accumulation assessment coefficient reference threshold, a normal change signal is generated, indicating that the electric field of the muscle tissue in the designated area is changing normally, and no risk warning is issued.

[0049] Once the deep learning model issues an accumulation risk warning, it automatically enters the real-time risk assessment mode and immediately uses the data provided by the micro field effect sensor to accurately calculate the real-time change value of the potential difference between different locations within the defined area and quantify the current severity of charge accumulation. The data provided by the micro field effect sensor is used to accurately calculate the real-time change value of the potential difference between different positions within the defined area and quantify the severity of the current charge accumulation. The specific steps are as follows: When the deep learning model issues an accumulation risk warning, it automatically enters the real-time risk assessment mode and uses the micro field effect sensor to measure the potential data at different locations in the designated area. The potential difference in the designated area is calculated by the instantaneous electric field change value collected by the micro field effect sensor, and the accumulation assessment coefficient is combined for quantitative analysis to generate a charge accumulation severity index to determine the severity of the charge accumulation. The calculation formula of the charge accumulation severity index is as follows: , in: is an indicator of the severity of charge accumulation, It is the accumulation assessment coefficient output by the deep learning model, which is used to measure the risk level of charge accumulation. is the reference threshold of the accumulation assessment coefficient, which is used to determine the critical point of the warning of charge accumulation. is the maximum potential difference at different positions within the demarcated area, indicating the degree of local polarization that may be caused by charge accumulation in the demarcated area. The calculation expression is: , in: and are the maximum and minimum electric field strengths in the demarcated area, respectively.

[0050] This step calculates the real-time potential difference and combines it with the accumulation evaluation coefficient for comprehensive processing to finally generate a charge accumulation severity index. If the charge accumulation severity index is too high, it indicates that the charge accumulation is serious and the electrical stimulation parameters need to be adjusted immediately.

[0051] After quantitatively calculating the potential difference, the pulse duty cycle adjustment action is automatically performed according to the intelligent control strategy. When the potential difference exceeds the preset potential difference threshold, the "low" state time of the pulse signal is extended, that is, the pulse gap is increased, the total amount of charge injection per unit time is reduced, and the degree of local accumulation is alleviated; The pulse duty cycle adjustment action is automatically performed according to the intelligent control strategy. The specific steps are as follows: Based on the calculated charge accumulation severity index , an adaptive pulse duty cycle control strategy is adopted to ensure uniform charge diffusion and avoid abnormal accumulation. The control expression is as follows: , in: is the adjusted pulse duty cycle used to reduce the effect of charge accumulation, is the reference pulse duty cycle, i.e. the default pulse duty cycle under normal circumstances. It is the adjustment coefficient that controls the sensitivity of duty cycle adjustment. A larger value means it is more sensitive to charge accumulation and has a larger adjustment range. when When the pulse duty cycle is adjusted Gradually reduce (i.e. extend the pulse interval and reduce the amount of charge injected per unit time) to promote the natural diffusion of charges and prevent damage to biological tissues caused by polarization effects or electrolytic reactions; when the charge accumulation severity index is reduced to close to the reference threshold of the accumulation assessment coefficient, the pulse duty cycle gradually returns to the baseline pulse duty cycle to ensure the normal function of muscle electrical stimulation.

[0052] Through the application of the above scheme, a significant beneficial effect is the realization of intelligent, real-time closed-loop control of the electrical stimulation process, which significantly improves the safety and effectiveness of treatment. This solution monitors the changes in the electric field in muscle tissue in real time, uses a deep learning model to identify the risk of charge accumulation in advance, and automatically adjusts the pulse duty cycle when the risk occurs, effectively preventing abnormal polarization and electrolytic reactions, and reducing the risk of tissue damage, cell necrosis and nerve damage. This system based on multi-level data processing and adaptive control can not only ensure that the charge is evenly diffused in the muscle, but also dynamically optimize the electrical stimulation parameters during the treatment process, making rehabilitation training and other electrical stimulation treatments more accurate and safe, and helping to improve the long-term rehabilitation effect of patients.

[0053] The present invention provides Figure 2 The electric shock simulation training detection and evaluation system shown includes an electric stimulation signal application module, an electric field monitoring module, a signal processing and feature extraction module, a deep learning risk prediction module, a real-time risk assessment module, and an adaptive pulse control module; The electrical stimulation signal application module applies a pre-set controllable electrical stimulation signal to the designated muscle area, simulating the electrical signal of the real external environment under precise control conditions; The electric field monitoring module directly monitors the electric field changes of muscle tissue in the defined area through micro field effect sensors, reflecting the diffusion trend of charge in real time; The signal processing and feature extraction module uses signal processing technology to pre-process the electric field change data acquired in real time, then extracts the key features reflecting charge accumulation, conducts in-depth analysis on the extracted key features, characterizes the risk of charge accumulation, and constructs a standardized feature vector for input into the deep learning model; The deep learning risk prediction module inputs the constructed feature vector into a pre-trained and fully verified deep learning model, analyzes the current real-time feature vector input, and outputs the potential charge accumulation risk prediction results in the charge diffusion trend, thus realizing early identification and early warning of charge accumulation risks; The real-time risk assessment module automatically enters the real-time risk assessment mode once the deep learning model issues an accumulation risk warning. It immediately uses the data provided by the micro field effect sensor to accurately calculate the real-time change value of the potential difference between different locations within the defined area and quantify the current severity of charge accumulation. After quantitatively calculating the potential difference, the adaptive pulse control module automatically performs pulse duty cycle adjustment according to the intelligent control strategy. When the potential difference exceeds the preset potential difference threshold, the "low" state time of the pulse signal is extended, that is, the pulse interval is increased, thereby reducing the total amount of charge injection per unit time and alleviating the degree of local accumulation.

[0054] The electric shock simulation training detection and evaluation method provided in the embodiment of the present invention is implemented by the above-mentioned electric shock simulation training detection and evaluation system. The specific methods and processes of the electric shock simulation training detection and evaluation system are detailed in the embodiment of the above-mentioned electric shock simulation training detection and evaluation method, which will not be repeated here.

[0055] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0056] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that 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 above 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.

[0057] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for detecting and evaluating electric shock simulation training, characterized in that: The following steps are involved: Apply a pre-set controllable electrical stimulation signal to the designated muscle area to simulate the electrical signal of the real external environment under precise control conditions; The electric field changes of muscle tissue in the defined area are directly monitored by micro field effect sensors, reflecting the diffusion trend of charge in real time. Signal processing technology is used to pre-process the electric field change data acquired in real time, and then the key features reflecting charge accumulation are extracted. The extracted key features are deeply analyzed to characterize the risk of charge accumulation, and a standardized feature vector is constructed for input into the deep learning model. The constructed feature vector is input into a pre-trained and fully verified deep learning model, the real-time feature vector currently input is analyzed, and the prediction result of potential charge accumulation risk in the charge diffusion trend is output to achieve early identification and early warning of charge accumulation risk; Once the deep learning model issues an accumulation risk warning, it automatically enters the real-time risk assessment mode and immediately uses the data provided by the micro field effect sensor to accurately calculate the real-time change value of the potential difference between different locations within the defined area and quantify the current severity of charge accumulation. After quantitatively calculating the potential difference, the pulse duty cycle adjustment action is automatically performed according to the intelligent control strategy. When the potential difference exceeds the preset potential difference threshold, the "low" state time of the pulse signal is extended, that is, the pulse interval is increased, the total amount of charge injection per unit time is reduced, and the degree of local accumulation is alleviated.

2. The electric shock simulation training detection and evaluation method according to claim 1, characterized in that: The demarcated area refers to the predetermined range of muscle or tissue that is directly acted upon by electrodes or affected by charge accumulation during muscle electrical stimulation experiments or treatments. It is precisely selected based on the anatomical structure, electrical stimulation target, clinical treatment needs or experimental design.

3. The electric shock simulation training detection and evaluation method according to claim 1, characterized in that: Features reflecting the charge accumulation situation are extracted, wherein the extracted features include the cumulative value of the time required for the potential to return to the baseline level and the cumulative value of the potential change rate. The time required for the potential to return to the baseline level and the cumulative value of the potential change rate are deeply analyzed to generate a potential regression delay reference value and a charge accumulation rate reference value, respectively. The charge accumulation risk is characterized by the potential regression delay reference value and the charge accumulation rate reference value, and a standardized feature vector is constructed for input into the deep learning model.

4. The electric shock simulation training detection and evaluation method according to claim 3, characterized in that: The constructed potential regression delay reference value and charge accumulation rate reference value are input into a pre-trained and fully verified deep learning model, the real-time feature vector of the current input is analyzed, and the accumulation assessment coefficient is output through the deep learning model. Based on the accumulation assessment coefficient, early identification and warning of potential charge accumulation risks in the charge diffusion trend are carried out.

5. The electric shock simulation training detection and evaluation method according to claim 4, characterized in that: The accumulation assessment coefficient generated when predicting the potential charge accumulation risk in the charge diffusion trend through the deep learning model is compared and analyzed with the preset accumulation assessment coefficient reference threshold to perform early identification and early warning of the charge accumulation risk. The specific steps are as follows: If the accumulation assessment coefficient is greater than the accumulation assessment coefficient reference threshold, a risk signal is generated, indicating that there is a risk of charge accumulation in the electric field change of the muscle tissue in the demarcated area, and an accumulation risk warning is issued; If the accumulation assessment coefficient is less than or equal to the accumulation assessment coefficient reference threshold, a normal change signal is generated, indicating that the electric field of the muscle tissue in the defined area is changing normally, and no risk warning is issued at this time.

6. The electric shock simulation training detection and evaluation method according to claim 5, characterized in that: The data provided by the micro field effect sensor is used to accurately calculate the real-time change value of the potential difference between different positions within the defined area and quantify the severity of the current charge accumulation. The specific steps are as follows: When the deep learning model issues an accumulation risk warning, it automatically enters the real-time risk assessment mode and uses the micro field effect sensor to measure the potential data at different locations in the designated area. The potential difference in the designated area is calculated by the instantaneous electric field change value collected by the micro field effect sensor, and the accumulation assessment coefficient is combined for quantitative analysis to generate a charge accumulation severity index to determine the severity of the charge accumulation. The calculation formula of the charge accumulation severity index is as follows: ， in: is an indicator of the severity of charge accumulation, It is the accumulation assessment coefficient output by the deep learning model, which is used to measure the risk level of charge accumulation. is the reference threshold of the accumulation evaluation coefficient, is the maximum potential difference at different positions within the demarcated area, indicating the degree of local polarization caused by charge accumulation in the demarcated area. The calculation expression is: ， in: and are the maximum and minimum electric field strengths in the demarcated area, respectively.

7. The electric shock simulation training detection and evaluation method according to claim 6, characterized in that: The pulse duty cycle adjustment action is automatically performed according to the intelligent control strategy. The specific steps are as follows: Based on the calculated charge accumulation severity index , an adaptive pulse duty cycle control strategy is adopted to ensure uniform charge diffusion. The control expression is as follows: ， in: is the adjusted pulse duty cycle used to reduce the effect of charge accumulation, is the reference pulse duty cycle, i.e. the default pulse duty cycle under normal circumstances. It is the adjustment coefficient that controls the sensitivity of duty cycle adjustment.

8. An electric shock simulation training detection and evaluation system, used to implement the electric shock simulation training detection and evaluation method according to any one of claims 1 to 7, characterized in that: It includes an electrical stimulation signal application module, an electric field monitoring module, a signal processing and feature extraction module, a deep learning risk prediction module, a real-time risk assessment module, and an adaptive pulse control module; The electrical stimulation signal application module applies a pre-set controllable electrical stimulation signal to the designated muscle area, simulating the electrical signal of the real external environment under precise control conditions; The electric field monitoring module directly monitors the electric field changes of muscle tissue in the defined area through micro field effect sensors, reflecting the diffusion trend of charge in real time; The signal processing and feature extraction module uses signal processing technology to pre-process the electric field change data acquired in real time, then extracts the key features reflecting charge accumulation, conducts in-depth analysis on the extracted key features, characterizes the risk of charge accumulation, and constructs a standardized feature vector for input into the deep learning model; The deep learning risk prediction module inputs the constructed feature vector into a pre-trained and fully verified deep learning model, analyzes the current real-time feature vector input, and outputs the potential charge accumulation risk prediction results in the charge diffusion trend, thus realizing early identification and early warning of charge accumulation risks; The real-time risk assessment module automatically enters the real-time risk assessment mode once the deep learning model issues an accumulation risk warning. It immediately uses the data provided by the micro field effect sensor to accurately calculate the real-time change value of the potential difference between different locations within the defined area and quantify the current severity of charge accumulation. After quantitatively calculating the potential difference, the adaptive pulse control module automatically performs pulse duty cycle adjustment according to the intelligent control strategy. When the potential difference exceeds the preset potential difference threshold, the "low" state time of the pulse signal is extended, that is, the pulse interval is increased, the total amount of charge injection per unit time is reduced, and the degree of local accumulation is alleviated.

9. An electric shock simulation training detection and evaluation applet, comprising the electric shock simulation training detection and evaluation system according to claim 8.

Images