Digital acupuncture dynamic control optimization method

Abstract

The invention relates to a digital acupuncture dynamic control optimization method. According to the technical scheme, the method comprises the steps that real-time physiological data, including the heart rate, the oxyhemoglobin saturation, the skin temperature and the muscle tension, of a patient are collected through a wearable device and a sensor; performing cleaning, denoising and feature extraction on the multi-modal physiological data by using a data preprocessing algorithm; a personalized digital acupuncture model is established, and the physical condition and acupoint sensitivity of a patient are dynamically evaluated; adjusting acupuncture output parameters, control frequency, pulse width and current intensity according to real-time feedback of a patient by adopting a self-adaptive fuzzy control algorithm through the obtained personalized digital acupuncture model; multi-channel biofeedback equipment including EEG, EMG and ECG is integrated, and instant response of a patient is monitored; through a closed-loop control algorithm, biological feedback data are input into the control model in real time, and dynamic adjustment of acupuncture and moxibustion parameters is controlled; a deep reinforcement learning algorithm is introduced into acupuncture therapy.

Description

Technical Field

[0001] The invention relates to a method for dynamic control optimization of acupuncture, and in particular to a method for dynamic control optimization of digital acupuncture. Background Art

[0002] The digital acupuncture dynamic control optimization technology currently available on the market, especially the method and system described in Chinese invention patent No. 2022116538540, has indeed promoted the intelligence of acupuncture treatment to a certain extent. However, through in-depth analysis of the actual application of this patented technology, we can find that it still has obvious shortcomings and limitations in many key aspects, which limit its wide application and efficacy improvement in actual medical environments. First of all, although this technology introduces dynamic tracking and acupoint recognition of the subject's body posture, it mainly relies on static image processing and body posture analysis technology, and matches it through an established acupoint database. However, since the physiological response of the human body is complex and dynamically changing, it is often difficult to capture the real-time physiological changes of individuals during treatment by analyzing only static images and existing acupoint locations. Therefore, the flexibility and real-time response ability of this method in personalized treatment are weak, and it is difficult to effectively respond to the physiological fluctuations and immediate feedback of patients during treatment.

[0003] Secondly, the patent emphasizes the monitoring and analysis of physiological signs and acupoint physiological characteristics, but its data collection and analysis are mainly based on preset thresholds and rules. This method lacks the in-depth understanding and dynamic adjustment capabilities of complex biofeedback signals to a certain extent. Traditional preset rules are often difficult to adapt to individual differences in patients, especially when facing patients with large changes in physiological state, which is prone to overtreatment or insufficient stimulation. In addition, although the patent mentions the use of execution process parameter data and environmental factor characteristics to infer and evaluate the patient's physiological state, in fact, the impact of environmental factors (such as temperature, humidity, background noise, etc.) on the efficacy of acupuncture is often indirect and complex. This comprehensive analysis reasoning mechanism relies more on fixed algorithms and lacks a data-driven adaptive learning mechanism. Therefore, when environmental conditions or patient physiological states suddenly change, the system's adjustment ability is relatively lagging, and it cannot respond to the patient's immediate needs in real time, which may lead to unstable treatment effects. Third, although this technology has the function of generating multi-dimensional acupuncture optimization strategies based on dynamic programming, the generation of its optimization strategies is more based on predefined rules and parameter adjustments, rather than real-time individualized data-driven decisions. This rule-driven strategy generation model often seems powerless when faced with complex and changing clinical situations. For example, the sensitivity of acupoints and tolerance to stimulation vary greatly among different patients. Simply relying on a fixed optimization strategy cannot flexibly adjust the intensity, frequency and coverage of acupuncture stimulation during treatment.

[0004] In addition, the patented method relies more on preset standardized data sets when dynamically evaluating and generating acupuncture effect evaluation reports, without fully considering the individual differences of different patients and the accumulation of long-term physiological feedback, which limits the room for improvement of its personalized treatment effect. More importantly, although the system can monitor and evaluate the physiological and emotional state of the subjects, it still has a large bottleneck in the speed of data processing and optimization feedback. It is based on the traditional rule reasoning method, and the efficiency of data processing and analysis is relatively low, making it difficult to achieve true real-time regulation. Therefore, in the actual treatment process, when the patient's physiological feedback signal fluctuates violently, the system may not be able to adjust the treatment parameters in time, which may increase the patient's discomfort or reduce the treatment effect. In addition, the feedback mechanism of the system is mainly based on preset parameter adjustments, and lacks the ability to deeply learn and self-optimize feedback data. This means that when the patient's physiological response fluctuates abnormally, it is difficult for the system to learn and adjust autonomously, which may miss the best time for treatment. Summary of the invention

[0005] The purpose of the present invention is to provide a method for dynamic control optimization of digital acupuncture, so as to solve some of the drawbacks and shortcomings pointed out in the background technology.

[0006] The present invention solves the above-mentioned technical problems by adopting the following technical solution, comprising the following steps:

[0007] S1. Analysis of digital acupuncture needs based on individual physiological signals:

[0008] S1.1. Collect patients’ real-time physiological data through wearable devices and sensors, including heart rate, blood oxygen saturation, skin temperature, and muscle tension;

[0009] S1.2. Use data preprocessing algorithms to clean, denoise, and extract features from multimodal physiological data;

[0010] S1.3. Establish a personalized digital acupuncture model to dynamically assess the patient's physical condition and acupoint sensitivity;

[0011] S2. Dynamic acupuncture stimulation control based on adaptive algorithm:

[0012] S2.1, the personalized digital acupuncture model obtained through S1.3, uses an adaptive fuzzy control algorithm to adjust the acupuncture output parameters according to the patient's real-time feedback, and controls the frequency, pulse width and current intensity;

[0013] S3. Dynamic optimization of closed-loop biofeedback control mechanism:

[0014] S3.1. Integrate multi-channel biofeedback equipment including EEG, EMG, and ECG to monitor the patient's immediate response;

[0015] S3.2, through the closed-loop control algorithm, the biofeedback data is input into the control model in real time to control the dynamic adjustment of acupuncture parameters;

[0016] S4. Dynamic control strategy iteration based on reinforcement learning:

[0017] S4.1. Introduce deep reinforcement learning algorithms in acupuncture treatment to build a dynamic evaluation and adjustment mechanism for treatment effects;

[0018] S4.2. Use the patient's biofeedback signal as a reward function and optimize the acupuncture parameter combination, including stimulation frequency, intensity, and duration, through the accumulation of multiple treatment sessions;

[0019] S5. Digital optimization of intelligent energy field control technology:

[0020] S5.1. Use a multimodal stimulation technique combining low-intensity electromagnetic pulses and infrared lasers to form a regulated energy field in the target acupuncture point area;

[0021] S5.2, through the digital energy field dynamic adjustment algorithm, according to the parameters obtained by biofeedback in step S4.2 and reinforcement learning in step S4.1, intelligently adjust the intensity and coverage of the energy field;

[0022] S5.3. Continuously monitor the temperature, blood flow and conductivity of the treatment area through sensors.

[0023] Furthermore, the establishment of a personalized digital acupuncture model includes:

[0024] The collected multimodal physiological data X(t) is input into the deep learning architecture integrating convolutional neural network (CNN) and recurrent neural network (RNN) to identify and predict the patient's acupoint sensitivity S a ; Define the following nonlinear sensitivity function:

[0025]

[0026] Among them, S a represents the predicted acupoint sensitivity, reflecting the patient's response intensity to acupuncture stimulation; X(t) is the multimodal physiological signal data at time t, including HRV, SpO2, skin conductance, and EMG; f 1 (X(t),θ 1 ) is the nonlinear function in the deep learning model, θ 1 represents the model weight parameter, obtained through training, which is used to capture the pattern of the patient's physiological signal; g 1 (w i ,X i ) is each feature X i The feature mapping function, wi is the weight coefficient of each feature.

[0027] Furthermore, the establishment of a personalized digital acupuncture model includes:

[0028] The integrated real-time biofeedback control system collects the patient's pain feedback, skin conductance changes and heart rate fluctuations in real time, and adjusts the acupuncture output parameter C(t) according to the feedback data; the closed-loop feedback regulation function is defined as follows:

[0029]

[0030] Where C(t) is the output control signal of the acupuncture device at time t, including stimulation frequency and intensity; h 1 (F(t),S a ) is the feedback adjustment function, which combines the current biofeedback F(t) and the predicted acupoint sensitivity S a Combination, used to adjust output parameters; is the time derivative of the feedback adjustment function, which indicates the immediate response capability to the feedback signal; λ is the adjustment coefficient, which is used to balance the immediate response of the feedback signal with the long-term cumulative effect; k 1 (F(τ)) is the cumulative impact factor of the historical feedback signal F(τ), which is used to adjust the effect of long-term stimulation.

[0031] Furthermore, the establishment of a personalized digital acupuncture model includes:

[0032] The reinforcement learning algorithm is introduced; the patient's biofeedback signal is used as the reward signal R(t) to guide the dynamic adjustment of the acupuncture parameter P(t); the treatment plan is optimized through the deep reinforcement learning model; the optimization target is described by the following formula:

[0033]

[0034] Among them, P(t) is the optimal stimulation parameter at time t, including the combination of frequency and intensity; α is the acupuncture stimulation frequency adjustment parameter, which controls the number of pulses per second; β is the acupuncture stimulation intensity adjustment parameter, which adjusts the current intensity of each pulse; R(t) is the reward signal, which is based on the patient's physiological feedback data; u(α,β,S a ,C(t)) is the comprehensive utility function, which is used to measure the effectiveness of treatment parameters and patient comfort.

[0035] Furthermore, the digital optimization method of the intelligent energy field control technology includes:

[0036] The sensor system is used to collect the patient's physiological feedback signals in real time, including skin temperature T(t), blood flow changes B(t) and local conductance E(t); the signal is preprocessed and feature extracted through the data processing module, and key features including change rate and fluctuation amplitude are extracted; the physiological signals are integrated to define the comprehensive feedback function F(t) to quantify the patient's overall physiological response:

[0037]

[0038] Among them, F(t) is a comprehensive feedback signal that reflects the overall changes in physiological parameters; a high F(t) value indicates a strong response of the patient to the energy field stimulation; p 1 (T(t), B(t), E(t)) is the instantaneous change function of the feedback signal, which captures the immediate physiological changes at the current moment t and is used for rapid response; T(t) is the skin temperature, reflecting the temperature change after local stimulation; B(t) is the blood flow change, which is used to evaluate the improvement of local blood circulation; E(t) is the local conductance, indicating the change of the electrical properties of the skin, which is related to the effect of acupuncture stimulation; q 1 (T(τ), B(τ), E(τ)) is the cumulative feedback effect function, which integrates the impact of historical data on the current state, and τ is the integral variable representing time; It represents the derivative with respect to time t and is used to capture the instantaneous change trend of the signal.

[0039] Furthermore, the digital optimization method of the intelligent energy field control technology includes:

[0040] Introduce an energy field intensity adjustment mechanism based on biofeedback; set the threshold θ of the feedback signal. When the comprehensive feedback F(t) exceeds the set threshold, the intensity I(t) of the energy field will be adjusted. The formula for adjusting the energy field intensity is:

[0041]

[0042] Where I(t) is the adjusted energy field intensity, which is used to control the actual treatment output; I 0 is the initial set energy field strength, which is the default basic strength; κ is the adjustment coefficient, which controls the sensitivity of the adjustment. The larger the value, the more sensitive it is to the feedback signal; θ is the threshold of the feedback signal, which indicates the ideal feedback level; when F(t)>θ, the system will reduce the intensity; otherwise, it will increase the intensity.

[0043] Furthermore, the digital optimization method of the intelligent energy field control technology includes:

[0044] The coverage of the energy field is dynamically adjusted through intelligent algorithms; the coverage area of ​​the energy field is adjusted by combining the sensor scanning data with the patient's individual anatomical model; when the feedback signal shows that the stimulation of a certain area is too strong or ineffective, the coverage is adjusted. The formula for adjusting the coverage is as follows:

[0045]

[0046] Among them, A(t) is the coverage of the energy field after adjustment; A 0 is the initially set coverage range, which is used to set the default action area of ​​the energy field; λ is the coverage adjustment coefficient, which is used to adjust the sensitivity of the energy field action area; To accumulate the absolute value integral of the feedback signal, the deviation of the feedback signal over a long period of time is evaluated to adjust the dynamic changes of the energy field coverage.

[0047] Beneficial effects of the present invention:

[0048] Traditional acupuncture treatment mainly relies on the experience of acupuncturists to select acupoints and adjust the intensity of stimulation. However, the present invention introduces a multimodal physiological signal acquisition system to monitor the patient's physiological feedback signals (such as skin temperature, blood flow changes and conductivity) in real time. These data are analyzed by intelligent algorithms to dynamically generate personalized treatment plans to ensure that acupuncture stimulation is performed in the patient's optimal physiological state. Therefore, the present invention greatly improves the accuracy and individualized effect of treatment, making acupuncture treatment more scientific and objective.

[0049] By combining deep learning and reinforcement learning algorithms, the intensity and coverage of the energy field can be automatically adjusted based on the patient's biofeedback signals. In traditional acupuncture treatment, the adjustment of acupuncture intensity and stimulation time usually depends on manual experience, while the present invention can automatically adjust according to the patient's physiological feedback through an intelligent closed-loop control system, realizing full automation and intelligence of the treatment process, reducing the impact of human errors.

[0050] Through the self-optimization capabilities of real-time biofeedback and reinforcement learning algorithms, the present invention can continuously optimize treatment parameters according to the patient's immediate physiological state and long-term feedback data, so that the treatment effect is gradually improved. In addition, the system can automatically adjust the output parameters of the energy field according to the changes in the treatment effect, so as to achieve the ideal treatment effect in a shorter time and improve the treatment efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is a flow chart of a method for dynamic control optimization of digital acupuncture according to the present invention.

[0052] Figure 2 A flow chart of the method for establishing a personalized digital acupuncture model for the present invention.

[0053] Figure 3 This is a flow chart of the digital optimization method of the intelligent energy field control technology of the present invention. DETAILED DESCRIPTION

[0054] The specific implementation modes of the present invention will be described in detail below in conjunction with the accompanying drawings.

[0055] Combined with Figure 1 As shown in FIG. 1 , a method for dynamic control optimization of digital acupuncture is first started through step S1, that is, digital acupuncture demand analysis based on individual physiological signals to start the entire treatment process. At this stage, through wearable devices and sensors worn on the patient, the system will collect a variety of physiological data in real time, including key parameters such as heart rate (HeartRate), blood oxygen saturation (SpO2), skin temperature (SkinTemperature) and muscle tension (MuscleTension). These data directly reflect the patient's current physiological state and physical response. Since physiological signals are often affected by environmental noise and equipment interference during the acquisition process, the system will first use a data preprocessing algorithm to clean and denoise these multimodal data to ensure the accuracy and effectiveness of the data. The data cleaning step includes filtering out invalid signals, processing lost data, and eliminating fluctuations caused by electromagnetic interference; denoising uses an adaptive filtering algorithm to remove high-frequency noise and abnormal fluctuations, thereby extracting high-quality basic data.

[0056] Next, the system will further analyze these processed physiological signals through feature extraction algorithms to extract the core features of the patient's current physiological state, such as heart rate variability (HRV), blood oxygen fluctuation rate, skin temperature gradient, and dynamic response of muscle tension. These feature data not only represent the patient's immediate physiological response, but also reveal the patient's sensitivity to external stimuli and overall health status. After completing the feature extraction, the system will use a deep learning model to establish a personalized digital acupuncture demand analysis model based on these individualized physiological parameters. Through machine learning algorithms (such as the combination of convolutional neural networks (CNN) and long short-term memory networks (LSTM), the system can learn and identify patients' unique physiological patterns from a large amount of historical physiological data and current real-time data, thereby dynamically predicting and evaluating the patient's physical condition and the sensitivity of acupoints. The establishment of this model enables the system to customize personalized treatment plans based on each patient's physiological data, rather than relying on the one-size-fits-all fixed acupoint and stimulation intensity settings in traditional acupuncture. By dynamically evaluating acupoint sensitivity, the system can identify the patient's optimal response time to acupuncture stimulation in different time periods, so as to more accurately select and adjust the location, intensity and frequency of acupoint stimulation, making the treatment effect more significant and the patient experience more comfortable.

[0057] Among them, step S2 is the dynamic acupuncture stimulation regulation based on the adaptive algorithm. After completing step S1 and establishing a personalized digital acupuncture model through S1.3, the system enters the dynamic regulation stage. The core of this stage is to use the previously obtained personalized model to fine-tune the acupuncture treatment parameters to adapt to the patient's physiological feedback and individual needs during the treatment process. First, the system will adopt an adaptive fuzzy control algorithm, which combines the advantages of fuzzy logic and adaptive control, and can achieve more flexible parameter adjustment in uncertain and nonlinear system environments. By inputting the patient's real-time physiological feedback signals, such as heart rate, blood oxygen saturation, skin conductance, and muscle response, the system can dynamically adjust the key output parameters of acupuncture stimulation, including stimulation frequency, pulse width, and current intensity. Specifically, the adaptive fuzzy control algorithm will obtain the patient's feedback data in real time and compare it with the baseline value in the personalized digital acupuncture model.

[0058] If a significant change in the patient's physiological feedback signal is detected, such as a sudden increase in heart rate, a significant increase in skin conductance, or abnormal fluctuations in muscle tension, the system will automatically adjust the output of the acupuncture equipment to ensure that the stimulation intensity and frequency always remain within the patient's acceptable range. Frequency regulation controls the depth of stimulation and frequency response by adjusting the number of pulses per second; pulse width adjustment is used to change the duration of each stimulation, thereby adjusting the depth of energy transfer; and current intensity regulation directly affects the patient's sensitivity to stimulation and overall therapeutic effect. Through this adaptive fuzzy control mechanism, the system is able to achieve instant response and adjustment during the treatment process without human intervention, greatly improving the personalization and effectiveness of the treatment. Compared with traditional fixed parameter acupuncture methods,

[0059] Step S3 is the dynamic optimization of the closed-loop biofeedback control mechanism. After establishing a personalized model and realizing adaptive regulation through steps S1 and S2 in the early stage, the system further introduces a closed-loop biofeedback control mechanism to ensure the accuracy and immediate responsiveness of the treatment process. Specifically, this step comprehensively monitors the patient's immediate physiological reactions during the treatment process by integrating multi-channel biofeedback equipment (including electroencephalogram (EEG), electromyogram (EMG) and electrocardiogram (ECG). EEG is used to detect the patient's brain activity state and can evaluate the patient's relaxation level or stress response under acupuncture stimulation; EMG is used to monitor muscle activity in real time, especially to capture changes in muscle tension and nervous system reactions; and ECG can provide real-time data on the patient's heart activity, such as heart rate and heart rate variability, to reflect the patient's overall state of relaxation and tolerance to stimulation.

[0060] At this stage, the system inputs the acquired biofeedback signals into the control model in real time through a closed-loop control algorithm, and dynamically adjusts the acupuncture stimulation parameters. The core of closed-loop control is to ensure that the system can quickly adjust according to the patient's immediate physiological feedback to achieve the best treatment effect. For example, if the EEG signal shows that the patient has a stress response under the current acupuncture stimulation, the system will automatically reduce the stimulation intensity or adjust the pulse frequency to reduce the patient's discomfort; if the EMG monitors that the muscle tension is too high, the pulse width and current intensity will be appropriately reduced to ensure that the stimulation is gentler for the patient; and when the ECG signal shows that the heart rate is too high, the system will immediately adjust the acupuncture parameters to avoid excessive load on the heart. This closed-loop biofeedback mechanism continuously optimizes the acupuncture parameters by continuously acquiring and analyzing biofeedback data, ensuring that the stimulation intensity, frequency and duration during the treatment process always match the patient's current physiological state, thereby improving the accuracy and individualization of the treatment. Through closed-loop control, the system can not only dynamically adjust the acupuncture output during the treatment process, but also gradually optimize the treatment strategy through the accumulation and analysis of historical data.

[0061] Wherein step S4 is the iteration of dynamic control strategy based on reinforcement learning. After completing the aforementioned personalized modeling, adaptive regulation and closed-loop biofeedback mechanism, the present invention further introduces a deep reinforcement learning algorithm to achieve self-optimization and continuous improvement of the acupuncture treatment process. At this stage, the system ensures that the various parameters of acupuncture treatment can be continuously adjusted and optimized as the patient's physiological response changes by constructing a dynamic evaluation and adjustment mechanism based on reinforcement learning. Specifically, the system first uses the patient's biofeedback signals during acupuncture treatment as the reward function input of the reinforcement learning algorithm, such as EEG, EMG and ECG signals. These feedback signals not only reflect the patient's immediate response to the current acupuncture stimulation, but also reveal the long-term impact of the treatment on his overall health status. Therefore, the design of the reward function not only considers the immediate physiological response (such as pain relief, muscle relaxation and heart rate reduction), but also considers the long-term efficacy of the treatment to ensure that the system can continuously optimize the combination of acupuncture parameters after each treatment session. Under the reinforcement learning framework, the system regards each treatment session as an independent learning round, and gradually optimizes the treatment strategy by continuously accumulating historical data. The system uses reinforcement learning algorithms such as Deep Q-Learning or Proximal Policy Optimization (PPO) to adjust the core parameters of acupuncture, including stimulation frequency, intensity and duration, according to the patient's biofeedback signals.

[0062] For example, when the patient's EEG signal shows an improvement in relaxation and the heart rate remains stable, the system will increase the reward value to encourage the current stimulation parameter settings; conversely, if the EMG signal shows an increase in muscle tension or the patient's heart rate fluctuations increase, the system will reduce the reward value and adjust the parameters to find a more suitable stimulation combination. This learning mechanism based on reward feedback enables the system to continuously optimize over multiple treatment rounds, thereby improving the efficiency and effectiveness of treatment. In addition, through the adaptive learning ability of the reinforcement learning algorithm, the system can autonomously explore the optimal acupuncture parameter combination when faced with individual differences among different patients, rather than relying on preset rules and fixed treatment plans.

[0063] Wherein step S5 is the digital optimization of intelligent energy field control technology. After completing personalized modeling, real-time feedback control and reinforcement learning strategy optimization in the previous steps, the present invention further introduces intelligent energy field control technology to maximize the efficacy of acupuncture treatment and improve the comfort of patients. At this stage, the system adopts a multimodal stimulation technology combining low-intensity electromagnetic pulses with infrared lasers to form a controllable energy field in the target acupuncture point area. The core advantage of this multimodal stimulation method is that it can act on different tissue levels at the same time, thereby achieving deeper and wider biological effects. Specifically, low-intensity electromagnetic pulses can stimulate changes in cell membrane potential, promote local blood circulation and cell metabolism, while infrared lasers increase the temperature of local tissues through heating effects, thereby accelerating blood flow and tissue repair. Therefore, the combined application of multimodal energy fields can achieve more significant therapeutic effects in a shorter time.

[0064] In order to ensure that the effect of the energy field can adapt to the patient's physiological changes in real time, the present invention adopts a digital energy field dynamic adjustment algorithm. The system intelligently regulates the intensity and coverage of the energy field by combining the real-time feedback data obtained from the biofeedback device (such as EEG, EMG, ECG) in step S4.2 and the treatment parameters optimized by the reinforcement learning algorithm in step S4.1. Specifically, the system dynamically adjusts the output intensity of the energy field according to the changes in the biofeedback signal. If during the treatment process, the EEG signal shows that the patient's stress response is aggravated, or the EMG signal detects an increase in muscle tension, the system will automatically reduce the intensity of the electromagnetic pulse and infrared laser to reduce the patient's discomfort. On the contrary, if the feedback signal shows that the patient's state tends to relax, the system may appropriately increase the energy output to improve the treatment effect. In addition, the system ensures that the stimulation is evenly distributed on the target acupuncture points and their surrounding tissues by adjusting the coverage of the energy field to avoid local discomfort caused by excessive concentration or insufficient treatment effect caused by energy dispersion.

[0065] To achieve the above-mentioned intelligent regulation, the system continuously monitors the temperature, blood flow rate and skin conductance of the target area through sensors during the treatment process. These monitoring data are not only used to adjust the output parameters of the energy field in real time, but also can be used as additional inputs to the reinforcement learning algorithm to further optimize the treatment strategy. For example, if the sensor detects that the temperature in the treatment area is too high, the system will reduce the intensity of the infrared laser to avoid burns to the patient's skin; if the blood flow sensor shows that the local blood circulation is not improved enough, the system will increase the frequency of electromagnetic pulses to enhance the blood flow effect. The conductivity data is used to evaluate the changes in the conductivity of the skin to ensure the effectiveness and safety of electromagnetic pulse stimulation during the treatment process. Through the fusion and analysis of multi-dimensional data, the system can make detailed fine-tuning of the energy field during the treatment process, making the treatment more accurate and highly personalized.

[0066] In summary, the S5 step achieves intelligent regulation of energy field intensity and coverage by utilizing multimodal stimulation technology of low-intensity electromagnetic pulses and infrared lasers, combined with a digital energy field dynamic regulation algorithm. This method not only significantly improves the effectiveness and safety of treatment, but also enables adaptive adjustments based on the patient's physiological feedback, ensuring the patient's comfort during treatment and continuous optimization of treatment effects.

[0067] Embodiment 1:

[0068] Combined with Figure 2 Flowchart, Mr. Zhang is a 45-year-old office worker who has been troubled by shoulder and neck pain and chronic fatigue for a long time. After a preliminary examination by the doctor, he was diagnosed with muscle tension syndrome caused by long-term sitting in the office. Although traditional acupuncture treatment helps relieve symptoms, the effect is not long-lasting. At the same time, Mr. Zhang feels uncomfortable with acupuncture, so the effect is limited. Based on this, the doctor suggested that he try a more accurate and personalized digital acupuncture dynamic control optimization method to improve the effect and reduce discomfort.

[0069] Before Mr. Zhang received treatment, his multimodal physiological data X(t) was first collected through wearable devices, including heart rate variability (HRV), blood oxygen saturation (SpO2), skin conductance, and electromyographic signals (EMG). In the preliminary test, the system showed that his HRV was 52ms, SpO2 was 96%, skin conductance was 23μS (micro-Siemens), and the peak value of the electromyographic signal was 45μV. These data reflect his current stress level and muscle tension. After the data collection is completed, the system uses a data preprocessing algorithm to filter out noise and standardize the signal to ensure the accuracy of subsequent data analysis.

[0070] The processed data X(t) is input into a deep learning architecture that integrates a convolutional neural network (CNN) and a recurrent neural network (RNN). CNN is used to extract the spatial features of multimodal physiological signals, while RNN is used to capture the temporal series dependencies of signals to identify the patient's response pattern to acupuncture stimulation. Through repeated training of historical data, the system learned Mr. Zhang's unique physiological pattern and predicted his acupoint sensitivity S a The system defines the following nonlinear sensitivity function to comprehensively evaluate Mr. Zhang’s physiological response:

[0071]

[0072] Formula explanation:

[0073] S a : The predicted acupoint sensitivity is used to reflect the intensity of Mr. Zhang's response to acupuncture stimulation. X(t): Multimodal physiological data at time t, including HRV, SpO2, skin conductance and EMG. 1 (X(t),θ 1 ): a nonlinear function used to capture the complex patterns of the patient’s physiological signals, θ 1 is the weight parameter of the deep learning model. 1 (w i ,X i ): Feature mapping function, based on different physiological characteristics X i Make adjustments, w i is the characteristic weight coefficient, which reflects the influence of each physiological signal on the sensitivity of acupoints.

[0074] The weight parameter θ obtained by the system through training 1 The value range is between [0.2, 0.9], and the feature weight coefficient w i The value range of is [0.1, 0.5], which is adjusted according to Mr. Zhang’s historical data. In the specific calculation, set:

[0075] f 1 (X(t),θ 1 )=0.5×HRV(t)+0.3×SpO2(t)+0.1×EMG(t)+0.1×skin conductance(t)

[0076] g 1 (w i ,X i )=0.4×HRV+0.2×SpO2+0.3×EMG+0.1×skin conductance

[0077] Taking Mr. Zhang's actual data (HRV=52ms, SpO2=96%, EMG=45μV, skin conductance=23μS), we can calculate:

[0078]

[0079]

[0080] Results a =117.4 indicates that Mr. Zhang is highly sensitive to acupuncture stimulation, so the system recommends choosing a lower current intensity and a shorter stimulation time during the initial treatment to ensure comfort.

[0081] The system passes S a The system analyzes Mr. Zhang's biofeedback data (such as the changing trends of HRV and EMG) and automatically adjusts the output parameters of the acupuncture equipment, including frequency, pulse width and current intensity. Considering Mr. Zhang's high sensitivity, the system sets the initial current intensity to 1.5mA (range 1.0-2.5mA), the frequency to 10Hz (range 5-20Hz), and the pulse width to 200μs (range 100-400μs). During the actual treatment process, the system automatically adjusts the output parameters by real-time monitoring of Mr. Zhang's biofeedback data (such as the changing trends of HRV and EMG) to ensure the treatment effect and patient comfort.

[0082] In the subsequent treatment sessions, the system will continue to optimize acupuncture parameters through reinforcement learning algorithms. The system uses Mr. Zhang's biofeedback as a reward signal and continuously searches for the optimal combination by adjusting the frequency, strength and duration. For example, after the second treatment, the system detected that Mr. Zhang's HRV increased to 58ms and his muscle tension decreased by 20%, indicating that the treatment was effective. Therefore, the system appropriately increased the current intensity to 2.0mA to further enhance the efficacy.

[0083] After four treatments, Mr. Zhang's shoulder and neck pain was significantly relieved, HRV was stable at more than 65ms, SpO2 remained above 98%, and muscle tension was reduced by 30%. After several treatments, Mr. Zhang's shoulder and neck pain had been significantly relieved, but in order to further consolidate the efficacy and improve the durability of the treatment, the doctor decided to use the closed-loop biofeedback control system in the digital acupuncture dynamic control optimization method of the present invention to continuously optimize the treatment parameters and respond more accurately to Mr. Zhang's physiological changes, thereby maximizing his health.

[0084] At this stage, by integrating the real-time biofeedback control system, the system will monitor and dynamically adjust Mr. Zhang's physiological signals more carefully.

[0085] First, the system monitors Mr. Zhang in real time during treatment through an integrated multi-channel biofeedback device. The specific monitored data include pain feedback F(t), skin conductance change E(t), and heart rate fluctuation HR(t). At the beginning of the latest treatment, Mr. Zhang's initial heart rate fluctuation was 70bpm (beats per minute), skin conductance was 22μS (micro-Siemens), and his subjective rating of pain F(t) was 6 (0-10 scale). These physiological data will continue to be collected throughout the treatment process to ensure that the system can respond to the patient's physiological changes in real time.

[0086] In order to better optimize the treatment parameters, the system uses a closed-loop feedback regulation function C(t) to dynamically adjust the output of the acupuncture device according to Mr. Zhang's real-time feedback signal. The output control signal of acupuncture stimulation is defined by the following formula:

[0087]

[0088] Formula explanation:

[0089] C(t): The output control signal of the acupuncture device at time t, including stimulation frequency and intensity. 1 (F(t),S a ): Feedback adjustment function, which feeds back F(t) and predicted acupoint sensitivity S in real time a Combined to optimize output parameters. The time derivative of the feedback adjustment function represents the ability to respond immediately to the feedback signal. This item is used to quickly adjust the output when the physiological signal changes suddenly. λ: Adjustment coefficient, used to balance the immediate response and cumulative effect, with a value range of [0.1, 0.5]. k 1 (F(τ)): Cumulative impact factor of historical feedback signals, taking into account the effect of long-term stimulation, τ is the time variable.

[0090] To further illustrate the actual operation of the system, take Mr. Zhang as an example. After the first few treatment sessions, the system has trained a personalized acupoint sensitivity S for Mr. Zhang. a =115, which indicates that he is more sensitive to acupuncture stimulation. In the latest treatment, the system began to collect his pain feedback F(t), with an initial value of 6. In order to calculate the closed-loop feedback control signal C(t), the actual data is substituted into the formula:

[0091] Set the feedback adjustment function to:

[0092] h 1 (F(t),S a )=0.6×F(t)+0.4×S a

[0093] therefore:

[0094] h 1 (6,115)=0.6×6+0.4×115=3.6+46=49.6

[0095] Next, calculate its time derivative:

[0096] Assuming that Mr. Zhang's pain feedback F(t) drops to 4 after 10 minutes, the change rate of the feedback adjustment function is:

[0097]

[0098] This suggests that the system detected a decrease in pain perception and therefore adjusted output parameters appropriately to maintain the therapeutic effect.

[0099] Next, the system adjusts the output parameters based on the patient's historical feedback signals, taking into account the long-term cumulative effect. The cumulative impact factor is set as:

[0100] k 1 (F(τ)) = 0.3 × F(τ)

[0101] Assuming that in the first 30 minutes of treatment, the average value of Mr. Zhang's pain feedback F(τ) is 5, the cumulative adjustment term is:

[0102]

[0103] Therefore, the final output control signal is:

[0104] C(t)=-0.2+0.3×45=-0.2+13.5=13.3

[0105] This means that at the current moment of treatment, the system will automatically adjust the output of the acupuncture device based on the feedback signal, adjusting the current intensity to a more appropriate level to ensure Mr. Zhang's treatment comfort.

[0106] Based on the above calculations, the system detects that Mr. Zhang's pain response is gradually weakening, so it will reduce the intensity of stimulation and slightly reduce the frequency to avoid overstimulation while maintaining the treatment effect. In subsequent treatment sessions, the system will further optimize the adjustment coefficient λ and the feedback function k through the reinforcement learning algorithm. 1 The system adjusts the weight of (F(τ)) to better adapt to Mr. Zhang’s physiological changes. For example, in subsequent treatments, if Mr. Zhang’s HRV and skin conductance show a higher level of relaxation, the system will gradually increase the current intensity and stimulation time to achieve a deeper therapeutic effect.

[0107] Embodiment 2:

[0108] Combined with Figure 3Flowchart, Ms. Zhang is a 52-year-old teacher who has long been troubled by lower limb pain and poor microcirculation, especially after standing for a long time. After the doctor's examination, it was determined that her problem was mainly caused by insufficient local blood circulation and muscle fatigue. The doctor decided to use the digital acupuncture dynamic control optimization method of the present invention, combined with intelligent energy field regulation technology for treatment, to improve lower limb microcirculation and relieve pain. The entire treatment plan optimizes energy field parameters through real-time biofeedback to ensure personalization and efficiency.

[0109] At the beginning of the treatment, Ms. Zhang was equipped with a sensor system to monitor her physiological feedback signals in real time. The monitored signals include skin temperature T(t), blood flow changes B(t), and local conductance E(t). Initially, the sensor detected the following data: skin temperature T(t) = 32.5°C, blood flow (per unit volume) B(t) = 20mL / min, and local conductance E(t) = 18μS. These data reflect Ms. Zhang's poor local circulation and poor vascular function. To ensure data accuracy, the system filters and standardizes the signal through a preprocessing module, removes environmental noise and abnormal data, and extracts key features such as temperature change rate, blood flow fluctuation amplitude, and conductivity change trend.

[0110] In order to quantify Ms. Zhang's overall physiological response to the treatment, the system integrates various physiological signals to define a comprehensive feedback function F(t) to capture instant changes and cumulative effects. The comprehensive feedback function expression is:

[0111]

[0112] Instantaneous change function p 1 (T(t), B(t), E(t)): used to capture instant changes in physiological signals, set as:

[0113] p 1 (T(t),B(t),E(t))=0.4·T(t)+0.3·B(t)+0.3·E(t)

[0114] In this formula, the weight coefficients 0.4, 0.3, and 0.3 reflect the importance of temperature, blood flow, and conductivity to the immediate response, respectively, and the value range is [0.1, 0.5].

[0115] Cumulative feedback effect function q 1 (T(τ), B(τ), E(τ)): Considering the impact of historical data on the current state, it is set as:

[0116] q 1 (T(τ),B(τ),E(τ))=0.5·ΔT(τ)+0.3·ΔB(τ)+0.2·ΔE(τ)

[0117] Among them, ΔT(τ), ΔB(τ), and ΔE(τ) represent the cumulative changes of temperature, blood flow, and conductance relative to the initial state, respectively.

[0118] According to the initial data T(0)=32.5℃, B(0)=20mL / min, E(0)=18μS, after 10 minutes of treatment, the sensor monitored that the temperature rose to T(10)=34.5℃, the blood flow increased to B(10)=28mL / min, and the conductance changed to E(10)=22μS.

[0119] Calculation of instantaneous change:

[0120] p 1 (T(10),B(10),E(10))=0.4·34.5+0.3·28+0.3·22=13.8+8.4+6.6=28.8

[0121]

[0122] Cumulative feedback term calculation:

[0123] q 1 (T(τ),B(τ),E(τ))=0.5·(34.532.5)+0.3·(2820)+0.2·(2218)

[0124] q 1 (T(τ),B(τ),E(τ))=0.5·2+0.3·8+0.2·4=1+2.4+0.8=4.2

[0125] Cumulative points are calculated as:

[0126]

[0127] Comprehensive feedback signal:

[0128] F(t)=0.47+42=42.47

[0129] Through calculation, Ms. Zhang's F(t) = 42.47 shows that her physiological state has a significant response to energy field stimulation. The system automatically adjusts the parameters of the energy field according to the feedback signal, increasing the electromagnetic pulse frequency from 10Hz to 12Hz, the infrared laser intensity from 1.2mW to 1.5mW, and the coverage range from the initial 5cm 2 Expand to 8cm 2 , to enhance the therapeutic effect and promote local microcirculation. After 5 treatments, Ms. Zhang's skin temperature stabilized at 35.5℃, blood flow increased to 32mL / min, and local conductance was maintained at 24μS. The patient reported that the pain score dropped from the initial 6 to 1, indicating that the pain was relieved and blood circulation was significantly improved.

[0130] In order to further optimize the treatment effect and ensure that the stimulation intensity is always within the most appropriate range, thereby improving the efficacy and reducing possible discomfort during the treatment, Ms. Zhang will monitor the comprehensive feedback F(t) in real time during the treatment and dynamically adjust the intensity I(t) of the energy field based on the set feedback signal threshold θ.

[0131] At the beginning of treatment, the system collected real-time data of Ms. Zhang's skin temperature T(t) = 34°C, blood flow change B(t) = 25mL / min, and local conductance E(t) = 20μS. These data were input into the comprehensive feedback function F(t) for calculation. 1 (T(t), B(t), E(t)) and the cumulative feedback effect function q 1 The settings of (T(τ), B(τ), E(τ)) are as follows:

[0132] p 1 (T(t),B(t),E(t))=0.5·T(t)+0.3·B(t)+0.2·E(t)

[0133] q 1 (T(τ),B(τ),E(τ))=0.4·(T(τ)-33)+0.4·(B(τ)-20)+0.2·(E(τ)-18)

[0134] The comprehensive feedback signal F(t) at the start of treatment is set to:

[0135]

[0136] The instantaneous change term is calculated as:

[0137] p 1 (Y(t),B(t),E(t))=0.5·34+0.3·25+0.2·20=17+7.5+4=28.5

[0138] Set the instantaneous change rate to 0.2 (unit: per minute) and the cumulative feedback integral to 25, then:

[0139] F(t)=0.2+25=25.2

[0140] The system sets the threshold value of the feedback signal to θ=22, which represents the ideal physiological feedback level.

[0141] According to the calculated value of the comprehensive feedback F(t), the system uses the following formula to adjust the energy field intensity I(t):

[0142]

[0143] in:

[0144] Initial strength I 0 =1.5mW;

[0145] Adjustment coefficient κ = 0.3 (value range is [0.1, 0.5]);

[0146] Threshold θ = 22;

[0147] The comprehensive feedback signal F(t)=25.2.

[0148] Substituting the data into the formula:

[0149]

[0150] I(t)=1.5·(1+0.3·0.145)=1.5·(1+0.0435)=1.5·1.0435=1.565mW

[0151] Therefore, the system increased the output intensity of the energy field to 1.565mW according to the feedback signal F(t) to enhance the local stimulation effect. As the treatment progressed, Ms. Zhang's blood flow improved significantly, and the sensor detected that F(t) increased to 28, exceeding the ideal threshold θ=22. At this time, the system automatically reduced the energy field intensity to avoid overstimulation. Substituting F(t)=28 into the formula:

[0152]

[0153] I(t)=1.5·(1+0.3·0.2727)=1.5·(1+0.08181)=1.5·1.08181=1.6227mW

[0154] The system reduced the energy field intensity to 1.6227mW to maintain Ms. Zhang’s physiological state within the ideal range and avoid discomfort caused by overstimulation.

[0155] As the treatment was coming to an end, the system detected that Ms. Zhang's comprehensive feedback signal dropped to F(t) = 20, which was lower than the threshold value θ = 22. At this time, the system decided to appropriately increase the energy field intensity to enhance the stimulation effect. Substituting F(t) = 20 into the formula:

[0156]

[0157] I(t)=1.5·(1-0.3·0.0909)=1.5·(1-0.02727)=1.5·0.97273=1.459mW

[0158] The system increases the intensity to 1.459mW, ensuring that the stimulation intensity is appropriately increased to maintain therapeutic effects.

[0159] Through the above-mentioned biofeedback-based intensity adjustment mechanism, the system dynamically adjusted the energy field intensity throughout the treatment process to ensure that the output intensity always matched Ms. Zhang's physiological state. After 6 treatments, Ms. Zhang's lower limb blood flow increased to 35mL / min, skin temperature reached 35.5℃, local conductance was maintained at 22μS, and pain score dropped to 0, indicating that the treatment effect was significant and there was no obvious discomfort.

[0160] Ms. Zhang hopes to further optimize the treatment to ensure that the coverage of the energy field can accurately act on the target area and avoid insufficient efficacy or discomfort caused by excessive concentration or dispersion. To this end, the system dynamically adjusts the coverage of the energy field through intelligent algorithms, combining sensor scanning data and Ms. Zhang's individual anatomical model to improve the accuracy of the treatment.

[0161] During this treatment, the system used sensors to monitor Ms. Zhang’s comprehensive feedback signal F(t) in real time, and set the initial coverage range A based on the patient’s individual anatomical model. 0 =5cm 2 At the beginning of the treatment, Ms. Zhang's skin temperature T(t) = 34.5°C, blood flow change B(t) = 28 mL / min, local conductance E(t) = 22 μS, initial value of the feedback signal F(t) = 23, feedback signal threshold θ = 22, coverage adjustment coefficient λ = 0.2 (value range [0.1, 0.5]).

[0162] The system dynamically adjusts the energy field coverage range according to the real-time changes of the feedback signal F(t) through the following formula:

[0163]

[0164] Formula parameter description:

[0165] A(t): Current energy field coverage (unit: cm 2 ).

[0166] A 0 =5cm 2 : Initial coverage.

[0167] λ=0.2: Coverage adjustment coefficient, which controls the sensitivity of coverage range changes.

[0168] The absolute value integral of the accumulated feedback signal is used to evaluate the accumulated deviation of the feedback signal.

[0169] During the treatment process, the changes of the comprehensive feedback signal F(t) are as follows:

[0170] F(5)=25, indicating that the feedback signal increases at 5 minutes;

[0171] F(10) = 24, indicating that the signal decreased slightly at 10 minutes;

[0172] F(15) = 20, indicating that the signal dropped below the threshold at 15 minutes;

[0173] F(20)=22, indicating that the signal returns to the ideal range.

[0174] Calculate the absolute value integral of the accumulated feedback signal:

[0175] In the first 5 minutes:

[0176]

[0177] In 5 to 10 minutes:

[0178]

[0179] In 10 to 15 minutes:

[0180]

[0181] In 15 to 20 minutes:

[0182]

[0183] Total accumulated points:

[0184]

[0185] Substitute the accumulated points into the coverage adjustment formula:

[0186] A(t)=5·(1+0.2·1.5908)=5·(1+0.31816)=5·1.31816=6.5908cm 2

[0187] The system adjusts the energy field coverage to A(t)=6.59cm 2 , ensuring that the stimulation range is expanded to act on peripheral areas where the feedback signal is abnormal.

[0188] At the end of the treatment, the system detected that Ms. Zhang's comprehensive feedback signal had returned to the threshold range, and the coverage range was automatically reduced to the initial value A. 0 =5cm 2 In actual operation, when the feedback signal exceeds the set threshold range, the system will immediately expand the coverage range; when the signal is below the threshold, the system will appropriately reduce the coverage range to avoid stimulation dispersion. The entire process realizes the dynamic adjustment of the energy field coverage range through the feedback loop to ensure the efficiency and accuracy of the treatment.

[0189] Through the intelligent energy field coverage range control technology, the system can accurately adjust the stimulation area according to Ms. Zhang's real-time feedback signal, avoiding discomfort caused by over-concentration or reduced efficacy due to excessive dispersion. After the treatment, Ms. Zhang reported that her lower limb pain completely disappeared, skin temperature remained at 35.8℃, blood flow increased to 32mL / min, and local conductance was maintained at 24μS. This treatment demonstrated the scientificity and efficiency of the digital acupuncture dynamic control optimization method in adjusting the coverage range, provided strong technical support for personalized treatment, and significantly improved the patient's treatment experience and efficacy.

[0190] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for dynamic control optimization of digital acupuncture, characterized in that The following steps are involved: S1. Analysis of digital acupuncture needs based on individual physiological signals: S1.

1. Collect patients’ real-time physiological data through wearable devices and sensors, including heart rate, blood oxygen saturation, skin temperature, and muscle tension; S1.

2. Use data preprocessing algorithms to clean, denoise, and extract features from multimodal physiological data; S1.

3. Establish a personalized digital acupuncture model to dynamically assess the patient's physical condition and acupoint sensitivity; S2. Dynamic acupuncture stimulation control based on adaptive algorithm: S2.1, the personalized digital acupuncture model obtained through S1.3, uses an adaptive fuzzy control algorithm to adjust the acupuncture output parameters according to the patient's real-time feedback, and controls the frequency, pulse width and current intensity; S3. Dynamic optimization of closed-loop biofeedback control mechanism: S3.

1. Integrate multi-channel biofeedback equipment including EEG, EMG, and ECG to monitor the patient's immediate response; S3.2, through the closed-loop control algorithm, the biofeedback data is input into the control model in real time to control the dynamic adjustment of acupuncture parameters; S4. Dynamic control strategy iteration based on reinforcement learning: S4.

1. Introduce deep reinforcement learning algorithms in acupuncture treatment to build a dynamic evaluation and adjustment mechanism for treatment effects; S4.

2. Use the patient's biofeedback signal as a reward function and optimize the acupuncture parameter combination, including stimulation frequency, intensity, and duration, through the accumulation of multiple treatment sessions; S5. Digital optimization of intelligent energy field control technology: S5.

1. Use a multimodal stimulation technique combining low-intensity electromagnetic pulses and infrared lasers to form a regulated energy field in the target acupuncture point area; S5.2, through the digital energy field dynamic adjustment algorithm, according to the parameters obtained by biofeedback in step S4.2 and reinforcement learning in step S4.1, intelligently adjust the intensity and coverage of the energy field; S5.

3. Continuously monitor the temperature, blood flow and conductivity of the treatment area through sensors.

2. A method for dynamic control optimization of digital acupuncture according to claim 1, characterized in that The establishment of a personalized digital acupuncture model includes: The collected multimodal physiological data X(t) is input into the deep learning architecture integrating convolutional neural network (CNN) and recurrent neural network (RNN) to identify and predict the patient's acupoint sensitivity S a ; Define the following nonlinear sensitivity function: Among them, S a represents the predicted acupoint sensitivity, reflecting the patient's response intensity to acupuncture stimulation; X(t) is the multimodal physiological signal data at time t, including HRV, SpO2, skin conductance, and EMG; f1(X(t),θ1) is the nonlinear function in the deep learning model, θ1 represents the model weight parameter, which is obtained through training and is used to capture the pattern of the patient's physiological signals; g1(w i ,X i ) is each feature X i The feature mapping function, w i is the weight coefficient of each feature.

3. A method for dynamic control optimization of digital acupuncture according to claim 2, characterized in that The establishment of a personalized digital acupuncture model includes: The integrated real-time biofeedback control system collects the patient's pain feedback, skin conductance changes and heart rate fluctuations in real time, and adjusts the acupuncture output parameter C(t) according to the feedback data; the closed-loop feedback regulation function is defined as follows: Where C(t) is the output control signal of the acupuncture device at time t, including stimulation frequency and intensity; h1(F(t), S a ) is the feedback adjustment function, which combines the current biofeedback F(t) and the predicted acupoint sensitivity S a Combination, used to adjust output parameters; is the time derivative of the feedback adjustment function, which represents the immediate response capability to the feedback signal; λ is the adjustment coefficient, which is used to balance the immediate response of the feedback signal with the long-term cumulative effect; k1(F(τ)) is the cumulative influence factor of the historical feedback signal F(τ), which is used to adjust the effect of long-term stimulation.

4. A method for dynamic control optimization of digital acupuncture according to claim 3, characterized in that The establishment of a personalized digital acupuncture model includes: introducing a reinforcement learning algorithm; using the patient's biofeedback signal as a reward signal to guide the dynamic adjustment of acupuncture parameters; and optimizing the treatment plan through a deep reinforcement learning model.

5. A method for dynamic control optimization of digital acupuncture according to claim 1, characterized in that The digital optimization method of the intelligent energy field control technology includes: using a sensor system to collect the patient's physiological feedback signals in real time, including skin temperature, blood flow changes and local conductance; preprocessing and feature extraction of the signals through a data processing module, and extracting key features including change rate and fluctuation amplitude.

6. A method for dynamic control optimization of digital acupuncture according to claim 5, characterized in that The digital optimization method of the intelligent energy field control technology includes: Introduce an energy field intensity adjustment mechanism based on biofeedback; set the threshold θ of the feedback signal. When the comprehensive feedback F(t) exceeds the set threshold, the intensity I(t) of the energy field will be adjusted. The formula for adjusting the energy field intensity is: Among them, I(t) is the adjusted energy field intensity, which is used to control the actual treatment output; I0 ​​is the initially set energy field intensity, which is the default basic intensity; κ is the adjustment coefficient, which controls the sensitivity of the adjustment. The larger the value, the more sensitive it is to the feedback signal; θ is the threshold of the feedback signal, which indicates the ideal feedback level; when F(t)>θ, the system will reduce the intensity; otherwise, it will increase the intensity.

7. A method for dynamic control optimization of digital acupuncture according to claim 6, characterized in that The digital optimization method of the intelligent energy field control technology includes: The coverage of the energy field is dynamically adjusted through intelligent algorithms; the coverage area of ​​the energy field is adjusted by combining the sensor scanning data with the patient's individual anatomical model; when the feedback signal shows that the stimulation of a certain area is too strong or ineffective, the coverage is adjusted. The formula for adjusting the coverage is as follows: Wherein, A(t) is the adjusted coverage range of the energy field; A0 is the initially set coverage range, which is used to set the default action area of ​​the energy field; λ is the coverage adjustment coefficient, which is used to adjust the sensitivity of the energy field action area; To accumulate the absolute value integral of the feedback signal, the deviation of the feedback signal over a long period of time is evaluated to adjust the dynamic changes of the energy field coverage.

Images