Transcranial magnetoacoustic stimulation closed-loop feedback method and system based on electroencephalogram guidance

Abstract

The invention provides a transcranial magnetoacoustic stimulation closed-loop feedback method and system based on electroencephalogram guidance. The method comprises the steps that original electroencephalogram signals of a target brain area are collected and preprocessed; target brain region activity features are extracted based on the preprocessed electroencephalogram signals, and the neural activity state of the target brain region is judged according to the target brain region activity features; according to the state judgment result, initial stimulation parameters are generated and optimized in combination with a preset stimulation response model; the generated initial stimulation parameters are converted into actual stimulation signals, the transcranial magnetoacoustic stimulation device is controlled to execute ultrasonic and magnetic field stimulation on the target brain area, and closed-loop feedback is conducted according to the activity state of the target brain area. According to the method, the limitation of a traditional transcranial stimulation method in the aspects of real-time feedback, individualized adaptation, artifact interference and safety control is effectively solved, accurate, controllable and individualized intervention of the neural activity of the target brain region is achieved, and a repeatable, safe and efficient technical means is provided for nerve regulation and control research and clinical application.

Description

Technical Field

[0001] This application belongs to the field of transcranial magnetic stimulation technology, and particularly relates to a closed-loop feedback method and system for transcranial magnetic acoustic stimulation based on electroencephalography (EEG). Background Technology

[0002] Electroencephalography (EEG), as a non-invasive neural signal monitoring technique, can be used to study brain functional states, neural oscillation characteristics, and their relationship with cognitive behavior. In recent years, non-invasive neuromodulation techniques such as transcranial magnetic stimulation (TMS) and transcranial ultrasound stimulation (TUS) have been widely used in neurorehabilitation, cognitive enhancement, and intervention for mental illnesses. Closed-loop neuromodulation technology, combining real-time neural signal monitoring and stimulation parameter adjustment, achieves dynamic regulation of neural activity in specific brain regions, becoming a cutting-edge research direction in neuroscience, rehabilitation medicine, and brain-computer interfaces.

[0003] CN118588174A discloses a method and apparatus for determining antigen-antibody docking information. The method includes: acquiring corresponding antigen-antibody docking information using a deep learning network based on antigen and antibody information. The antigen information includes antigen sequence information and antigen structure information, and the antibody information includes antibody sequence information and antibody structure information. The deep learning network includes an encoder-decoder architecture, and the antigen-antibody docking information includes antigen-antibody complex structure information. This deep learning network predicts the binding structure of unbound antigens and antibodies, predicting not only the structure of antigen-antibody binding but also the structural transformation of antigens and antibodies from the unbound to the bound state. Using an encoder-decoder architecture for binding structure prediction effectively improves computational efficiency and prediction accuracy.

[0004] The above scheme mainly uses deep learning to predict antigen / structure and antibody / structure information. It is a static prediction model that cannot respond to changes in experiments or in vivo feedback in real time and lacks a real-time monitoring and adjustment mechanism. When there is a deviation between the input antigen / antibody state and the experimental environment, the complex structure predicted by the model may be inaccurate. Summary of the Invention

[0005] In view of this, this application aims to propose a closed-loop feedback method and system for transcranial magnetoacoustic stimulation based on electroencephalography (EEG) to solve the problems of static prediction, lack of real-time feedback, insufficient adaptive ability, and large interference from experimental noise in traditional antigen-antibody prediction or neural stimulation intervention.

[0006] To achieve the above objectives, the technical solution of this application is implemented as follows: In a first aspect, this application provides a closed-loop feedback method for transcranial magnetoacoustic stimulation based on electroencephalography (EEG), comprising: Raw electroencephalogram (EEG) signals from the target brain region are acquired and preprocessed; wherein, the signal preprocessing includes bandpass filtering, adaptive noise cancellation, and artifact removal; Based on the preprocessed EEG signals, the activity characteristics of the target brain region are extracted, and the neural activity state of the target brain region is determined according to the activity characteristics of the target brain region. Among them, the power ratio of the target brain region is used as the core determination index. The phase difference and amplitude fluctuation range of the neural signals between the target brain region and the non-target brain region are calculated to dynamically evaluate the temporal synchronization and intensity changes of neural activity. Based on the state determination results, the initial stimulus parameters are generated and optimized in conjunction with the preset stimulus response model. The generated initial stimulation parameters are converted into actual stimulation signals. Ultrasonic and magnetic field stimulation is performed on the target brain region by controlling the transcranial magnetoacoustic stimulation device, and closed-loop feedback is performed based on the activity status of the target brain region.

[0007] Secondly, based on the same inventive concept, this application also provides a closed-loop feedback system for transcranial magnetoacoustic stimulation guided by electroencephalography (EEG), comprising: The signal acquisition module is configured to acquire raw electroencephalogram (EEG) signals from the target brain region and perform signal preprocessing; wherein, the signal preprocessing includes bandpass filtering, adaptive noise cancellation, and artifact removal; The feature extraction module is configured to extract activity features of the target brain region based on the preprocessed EEG signal, and determine the neural activity state of the target brain region based on the activity features of the target brain region; wherein, the power ratio of the target brain region is used as the core determination index, and the phase difference and amplitude fluctuation range of the neural signals between the target brain region and the non-target brain region are calculated to dynamically evaluate the temporal synchronization and intensity changes of neural activity. The parameter generation and optimization module is configured to generate and optimize initial stimulus parameters based on the state determination results and a preset stimulus response model. The magnetoacoustic stimulation execution module is configured to convert the generated initial stimulation parameters into actual stimulation signals, control the transcranial magnetoacoustic stimulation device to perform ultrasound and magnetic field stimulation on the target brain region, and provide closed-loop feedback based on the activity state of the target brain region.

[0008] Thirdly, based on the same inventive concept, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method described in the first aspect.

[0009] Fourthly, based on the same inventive concept, this application also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions for causing the computer to perform the method as described in the first aspect.

[0010] Compared with existing technologies, the EEG-guided transcranial magnetic acoustic stimulation closed-loop feedback method and system described in this application have the following advantages: The EEG-guided transcranial magnetoacoustic stimulation closed-loop feedback method described in this application achieves integrated control of the entire process from EEG signal acquisition, preprocessing, feature extraction to closed-loop stimulation and adaptive parameter optimization. Through high-resolution, multi-mode acquisition, dynamic threshold determination, and adaptive adjustment of safety constraints, it effectively solves the limitations of traditional transcranial stimulation methods in real-time feedback, individualized adaptation, artifact interference, and safety control. It achieves precise, controllable, and personalized intervention of neural activity in target brain regions, providing a repeatable, safe, and efficient technical means for neuromodulation research and clinical applications. Attached Figure Description

[0011] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of a closed-loop feedback method for transcranial magnetoacoustic stimulation based on electroencephalography (EEG) as described in an embodiment of this application. Figure 2 This is a timing diagram of the transcranial magnetoacoustic stimulation closed-loop feedback system described in the embodiments of this application; Figure 3 This is a flowchart illustrating the EEG acquisition and preprocessing process described in the embodiments of this application. Figure 4 This is a flowchart illustrating the closed-loop feedback and dynamic adjustment process described in the embodiments of this application. Figure 5 This is a schematic diagram of the structure of a closed-loop feedback system for transcranial magnetoacoustic stimulation based on electroencephalography (EEG) as described in an embodiment of this application. Figure 6 This is a schematic diagram of the hardware structure of the electronic device described in an embodiment of this application. Detailed Implementation

[0012] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with specific embodiments and the accompanying drawings.

[0013] It should be noted that, unless otherwise defined, the technical or scientific terms used in the embodiments of this application should have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms "first," "second," and similar terms used in the embodiments of this application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are only used to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0014] As described in the background section above, traditional methods lack closed-loop feedback and are limited by static prediction. This embodiment addresses the problems of static prediction, lack of real-time feedback, insufficient adaptive capability, and significant interference from experimental noise in traditional antigen-antibody prediction or neural stimulation interventions by proposing a closed-loop feedback method for transcranial magnetic acoustic stimulation based on electroencephalography (EEG). Through multi-channel real-time EEG monitoring, time-frequency feature extraction, dynamic state determination, and adaptive stimulation parameter optimization, precise control of neural activity in the target brain region is achieved. This ensures both the safety and individualization of the intervention while improving the real-time performance, stability, and effectiveness of the stimulation intervention, thus overcoming the shortcomings of existing technologies in real-time control accuracy and closed-loop adaptive capability.

[0015] The embodiments of this application are described in detail below with reference to the accompanying drawings.

[0016] Please see Figure 1 and Figure 2 As shown, this embodiment provides a closed-loop feedback method for transcranial magnetoacoustic stimulation based on electroencephalography (EEG), which specifically includes the following steps: Step S101: Acquire raw EEG signals from the target brain region and perform signal preprocessing; wherein, signal preprocessing includes bandpass filtering, adaptive noise cancellation, and artifact removal.

[0017] Specifically, in this embodiment, a multi-channel EEG electrode array is attached to the scalp surface of the subject. The layout of the electrode array is based on the international 10-20 system or its extended version (such as the 10-10 system) to cover the target brain region (such as the prefrontal cortex, parietal lobe, or temporal lobe). The electrode array layout is further defined as follows: a main electrode is positioned at the center of the target brain region, with at least four auxiliary electrodes symmetrically distributed around the main electrode, and the distance between the auxiliary electrodes and the main electrode is [missing information]. To ensure high-resolution acquisition of transient artifacts during routine EEG monitoring and stimulation, the sampling system needs to support two recording modes and be able to switch between or acquire data in parallel during the experimental process. The routine recording mode aims for long-term / steady-state EEG monitoring and real-time feature extraction, with a sampling rate of [missing information]. This ensures full coverage of the 0.5–100Hz effective frequency band and provides a margin for anti-aliasing and time-frequency analysis. The analog front-end should be equipped with an anti-aliasing analog low-pass filter with a cutoff frequency of [missing information]. High-frequency aliasing interference is suppressed by setting the frequency to approximately 450 Hz (500 Hz below the Nyquist frequency).

[0018] Stimulation-synchronous / High-sample mode: During ultrasound / magnetic stimulation, to accurately record the stimulation transients, rapid rise times, and high-frequency artifacts (and their spectral extensions), the sampling rate is increased. The corresponding analog anti-aliasing filter cutoff frequency This mode can be used for stimulus artifact modeling, template generation and post-processing (e.g., template subtraction, ICA, NLMS, etc.).

[0019] Furthermore, all channels should be sampled simultaneously to avoid phase deviation caused by sampling delays between channels; each channel's ADC needs to meet the following front-end performance indicators to ensure... Level EEG and ability to distinguish stimulus artifacts: The input impedance should be ≥24 bits (or equivalent effective bits ENOB to guarantee sub-μV input resolution), channel input impedance ≥100MΩ, common-mode rejection ratio (CMRR) ≥100dB, input-referred noise <0.5μVRMS (within the target bandwidth), and support for external trigger / TTL markers to synchronize stimulation pulses (timestamp accuracy ≤1μs). The front end should also include configurable hardware amplification / saturation protection and fast recovery mechanisms to prevent the amplifier from saturating for extended periods due to stimulation pulses.

[0020] The main electrode and auxiliary electrode together form a local spatial filtering network. A weighted average algorithm is used to calculate the comprehensive neural activity characteristics of the target brain region, with weighting coefficients... Distance between each electrode and the projection point of the target brain region cortex (Unit: cm) shows an inverse relationship, that is ; In the formula, It is a constant.

[0021] In practical implementation, the weights should be jointly verified with spatial filtering based on the conductor model or Laplace / lead-field inversion to quantify the bias caused by the inverse distance approximation.

[0022] After acquiring the raw EEG signals, a bandpass filter was used for preprocessing. The cutoff frequency range of the filter was 0.5Hz to 100Hz to remove environmental noise and electromyographic interference, while retaining the effective frequency bands related to neural activity. Furthermore, an adaptive noise cancellation algorithm was used to separate and suppress transient electromagnetic interference signals generated during the operation of the transcranial magnetic stimulation (TMS) device. The main frequency range of the interference signals was 100Hz to 500Hz, and their intensity was related to the peak current of the stimulation pulse. (Unit: A) Pulse Width (Unit: ms), and the spatial geometry between the electrodes and the stimulation coil / transducer are also relevant. For engineering simplification, the interference intensity can be initially approximated as an empirical calibration model: ; in, Units are , The interference coefficient characteristic of the equipment is obtained through experimental calibration. The experimental calibration process includes: under different... , Stimulus artifact data are collected under certain conditions, and multivariate regression or least squares fitting is used to obtain... And its confidence interval is obtained through cross-validation. For higher accuracy, the model is extended to: ; in, The distance between the electrode and the stimulus source (cm). The relative geometric parameters of the coil / transducer and electrodes are given. This functional form is decomposed into multiple coefficients through experimental fitting to capture the contribution of different factors to the interference.

[0023] The recommended artifact removal strategy is a hierarchical combination: hardware priority (masking, differential amplification, fast amplifier protection) → triggered synchronous blanking (short-term masking) → template subtraction → adaptive filtering (NLMS / RLS) → ICA / ASR / beamforming. The system should record the application timing and performance metrics of each method for offline evaluation.

[0024] Furthermore, the adaptive noise cancellation algorithm in step S101 employs the minimum mean square error (LMS) algorithm, and its update formula is as follows: ; in, For the first The filter weight vector of the next iteration Step size factor ( ), The error value between the preprocessed signal and the interference signal. The LMS step size is the input vector of the interference signal. The choice of must satisfy the theoretical convergence condition, for example: ; in, The input autocorrelation matrix is ​​the largest eigenvalue, and normalized LMS (NLMS) or RLS with a forgetting factor is considered in the implementation as a more robust alternative; the initial step size is recommended to be... The range is automatically scaled based on input power; By adjusting the weight vector in real time, the transient electromagnetic interference signal (frequency range 100Hz-500Hz) generated by the transcranial magnetic acoustic stimulation device is separated and subtracted from the original EEG signal, thereby reducing the interference intensity. The noise level is reduced to below the effective component of the EEG signal (e.g., below 0.5 μV) to ensure high-fidelity acquisition of neural activity characteristics of the target brain region.

[0025] This embodiment allows for switching or parallel acquisition during experiments by setting a regular recording mode (≥1kHz) and a stimulus-synchronized high sampling mode (≥5kHz), taking into account both steady-state monitoring and transient artifact capture. The high sampling mode is used to model stimulus artifacts and generate templates, providing a data foundation for adaptive filtering and post-processing, and achieving accurate artifact elimination.

[0026] Step S102: Extract the activity features of the target brain region based on the preprocessed EEG signals, and determine the neural activity state of the target brain region based on the activity features of the target brain region; wherein, the power ratio of the target brain region is used as the core judgment index, and the phase difference and amplitude fluctuation range of the neural signals between the target brain region and the non-target brain region are calculated to dynamically evaluate the temporal synchronization and intensity changes of neural activity.

[0027] Specifically, in this embodiment, such as Figure 3 As shown, by performing time-frequency analysis on the preprocessed EEG signals, short-time Fourier transform or continuous wavelet transform is used to extract the power spectral density of the target brain region in specific frequency bands (such as theta waves 4-8Hz, gamma waves 30-100Hz). and And calculate their ratio. ;when Value exceeds preset threshold When abnormal neural oscillatory activity is detected in the target brain region, stimulation intervention needs to be initiated; simultaneously, the phase difference of the neural signal is calculated based on a sliding time window. (Unit: radians) and amplitude fluctuation range (unit: This allows for the dynamic assessment of the temporal synchronicity and intensity changes of neural activity.

[0028] This step involves calculating the phase difference between the target brain region and the non-target brain region. This allows for the identification of cross-regional abnormal synchronization and adds inhibitory weights to non-target brain regions in the stimulation parameters. The innovation lies in the fact that closed-loop intervention not only focuses on the target brain region but can also regulate network-level neural synchronization activities, thereby improving the overall accuracy of the intervention and the neural network regulation capability.

[0029] The extraction of activity features of the target brain region specifically involves using dual-band power ratios. Among the core judgment indicators, The frequency band is 4-8Hz. The wavelength range is 30-100Hz; simultaneously, the phase difference of neural signals between the target brain region and non-target brain regions (such as the corresponding region in the contralateral hemisphere) is calculated. To determine if there is any abnormal synchronization across regions; if The absolute value of continuous Each time window (e.g., M=2) is less than If the signal is abnormally synchronized across regions, it is necessary to add inhibitory control weights for non-target brain regions to the stimulation parameters to achieve intervention for abnormal cross-regional neural synchronization. At the same time, the dual sampling mode (conventional sampling ≥1kHz, stimulation synchronization high sampling ≥5kHz) can accurately capture transient stimulus artifacts, providing a data basis for template generation and adaptive filtering, thereby improving the accuracy of closed-loop stimulation and the control effect at the neural network level.

[0030] This embodiment utilizes real-time EEG signal characteristics (power ratio) Phase difference Amplitude variation The system can dynamically determine the state of the target brain region and achieve closed-loop feedback regulation. Transcranial magnetoacoustic stimulation (TMS + ultrasound) can automatically adjust stimulation parameters according to real-time neural activity to ensure individualization and adaptability.

[0031] Step S103: Based on the state determination result, generate and optimize the initial stimulus parameters in conjunction with the preset stimulus response model.

[0032] Specifically, in this embodiment, based on the state determination result and in conjunction with a preset stimulus response model, initial stimulus parameters, including the ultrasonic pulse intensity, are generated. (Unit: W / cm) 2 ), static magnetic field strength (Unit: T) and stimulation frequency (Unit: Hz). For ease of implementation and interpretation, the following relationship is used as the initial control model (linear approximation) for the controller: ; ; ; in, These are the baseline values, The above linear relationship serves only as an initial control model. Actual neural responses are often nonlinear and vary greatly from person to person. Therefore, the system needs to introduce adaptive control mechanisms, such as Model Reference Adaptive Control (MRAC) or online parameter adjustment processes based on reinforcement learning or Bayesian optimization, to dynamically correct the model and improve robustness during closed-loop operation.

[0033] Furthermore, all control parameters must be calibrated in offline experiments and early clinical trials to determine their individualized baseline values, confidence intervals, and safety ranges. For example, The values ​​of the stimulus parameters need to have clearly defined upper and lower limits, and out-of-bounds protection logic should be set in the control loop to prevent the stimulus parameters from exceeding the safety boundary due to model bias or noise. Through this three-level mechanism of "linear approximation + online adaptation + safety boundary restriction", the individualization and stability of the closed-loop stimulus are improved while ensuring safety.

[0034] When optimizing stimulation parameters, the intensity of the ultrasound pulse adjustment coefficient It is related to the neuronal excitability threshold of the target brain region; specifically, the neuronal excitability threshold of different subjects was obtained through offline experiments. (unit: ), and establish The linear relationship model, in which and To fit the parameters and achieve personalized adjustment; static magnetic field strength Gain coefficient magnetic susceptibility parameters of the target brain region (unit: Related to this, calibrated through experiments. back, ,in, For system calibration coefficients; Online learning algorithms (RL, Bayesian optimization, or MRAC) must be subject to safety constraints during the exploration phase: safe RL or box constraints should be used to limit the action space; all candidate parameters should be evaluated by a safety check module before execution (including instantaneous energy, peak threshold, and cumulative dose), and conservative strategies or fallback to the most recently validated parameter set should be used if necessary.

[0035] This step employs a linear initial control model (ultrasonic intensity). static magnetic field Stimulation frequency The innovation lies in combining offline-calibrated individualized parameters (neuronal excitability threshold, magnetic susceptibility) and adjusting them in real time in the closed loop through MRAC / reinforcement learning / Bayesian optimization. The innovation is to combine linear models with online adaptive control to achieve individualized, safe and dynamically optimizable stimulation strategies.

[0036] Based on the power ratio of target brain regions Phase difference and amplitude change By determining the neural state in real time and combining an initial linear model with online adaptive control (MRAC, reinforcement learning, or Bayesian optimization), stimulation parameters can be dynamically adjusted to achieve individualized neural modulation. At the same time, the stimulation process is ensured to be safe and reliable through safety boundary constraints, instantaneous energy, and peak threshold limits.

[0037] Step S104: Convert the generated initial stimulation parameters into actual stimulation signals, and perform ultrasound and magnetic field stimulation on the target brain region by controlling the transcranial magnetoacoustic stimulation device, and perform closed-loop feedback according to the activity state of the target brain region.

[0038] Specifically, in this embodiment, the generated [ultrasound] is stimulated by an integrated ultrasonic transducer and a permanent magnet. and The parameters are converted into actual stimulus signals; the ultrasonic transducer emits ultrasonic waves in pulse form, with a pulse repetition frequency of [missing value]. Single pulse duration =0.1ms to 1ms adjustable, the static magnetic field is fixedly applied to the target brain region through a permanent magnet; during the stimulation process, real-time EEG signals are continuously collected through the electrode array, and the preprocessing and feature extraction processes of steps S101 and S102 are repeated to form a closed-loop feedback real-time data stream.

[0039] The hardware of the stimulation device includes: an electromagnetic shielding layer between the ultrasonic transducer and the permanent magnet, the shielding layer being made of... - Metal or permalloy material, thickness of =1-3mm, to limit the transient electromagnetic interference generated by the magnetoacoustic stimulation device to outside the detection frequency band (0.5Hz-100Hz) of the electrode array; at the same time, the conductors of the electrode array use twisted-pair shielded cables with a shielding efficiency of 1-3mm. (unit: )satisfy To ensure the signal-to-noise ratio of the acquired signal. (unit: Not less than .

[0040] The system must meet the relevant safety standards before any online or offline trials: ultrasound parameters should be based on diagnostic / research reference standards (e.g., mechanical index MI, ISPTA / ISPPA) and maximum permissible values ​​should be set; static magnetic field strength must comply with ICNIRP or relevant national public / occupational exposure limits. All sessions must record and monitor local temperature rise (e.g., probe / scalp), and maximum cumulative energy and maximum duration (e.g., maximum duration per session, session intervals) should be set. Ethical and regulatory approval documents and a safety assessment report must be submitted before human trials.

[0041] like Figure 4 As shown, the real-time extracted value, and Compared with the preset dynamic adjustment threshold, if Value continuous None of the time windows (e.g., N=3) exceeded ,and fluctuation range (For example )Stablize, absolute rate of change If the value is lower than the set value, it is determined that the activity in the target brain region has returned to normal, and stimulation is terminated; otherwise, based on the current... value, and The deviation needs to be recalculated. , and The parameters are updated, the stimulus signal is maintained, and the intervention continues until the termination condition is met.

[0042] When in real time Value deviates from preset threshold At that time, ultrasonic intensity Adjustment amount: ; in, This is a dynamic adjustment coefficient; Static magnetic field strength Adjustment amount: ; in, Benchmark for the target brain region Wave power, This refers to the dynamic gain coefficient. Stimulation frequency Adjustment amount: ; in, The baseline phase difference for the target brain region. This is the frequency adjustment coefficient; The criteria for terminating the intervention are further limited to: when the target brain region... Value continuous Each time window (e.g., N=3) is less than ,and fluctuation range and absolute rate of change When all stable thresholds are met, the stimulation device automatically enters a low-power standby mode, while simultaneously recording EEG data from the intervention process as a basis for subsequent personalized parameter optimization; if subsequent detection... The value exceeded again Then, based on historical data, the parameters can be quickly restored to the most recent effective stimulus without re-initialization.

[0043] This embodiment utilizes a multi-channel EEG electrode array for high-precision acquisition of target brain regions. By combining bandpass filtering, minimum mean square error (LMS / NLMS / RLS) adaptive noise cancellation algorithms, and multi-level artifact removal strategies, it achieves real-time suppression of transient electromagnetic interference generated by transcranial magnetoacoustic stimulation (TMS), controlling the interference intensity below the effective EEG signal (e.g., <0.5 μV) to ensure reliable signal acquisition. The electrode layout is based on the international 10-20 or 10-10 system, with main electrodes and symmetrical auxiliary electrodes positioned around the target brain region. Comprehensive neural activity characteristics are calculated using weighted averaging and spatial filtering networks, with weights inversely proportional to the distance from the electrode to the cortical projection point. This can be jointly validated with a conductor model or surface Laplace correction to achieve accurate quantification of local brain region activity. This method supports both conventional recording modes and stimulation-synchronized high-sampling modes. Synchronized or phase-locked ADCs avoid channel phase deviations and provide high resolution, low noise, and external trigger synchronization capabilities to meet the acquisition requirements of transient stimulation signals.

[0044] In the signal processing and feature extraction stages, this embodiment obtains the power spectral density of theta waves (4-8Hz) and gamma waves (30-100Hz) of the target brain region through short-time Fourier transform or continuous wavelet transform, and calculates the dual-band power ratio. and phase difference and amplitude fluctuation This is used to identify abnormal neural oscillations and cross-regional synchronization abnormalities, thereby determining whether to initiate stimulation intervention and adding inhibitory weights for non-target brain regions to the stimulation parameters. The stimulation parameters are generated based on an initial control model with a linear approximation, including ultrasound intensity. static magnetic field Stimulation frequency ,pass , and Initial adjustments are made, and closed-loop online adaptive corrections are achieved by combining model reference adaptive control (MRAC), reinforcement learning, or Bayesian optimization to address the nonlinearity of neural responses and individual differences.

[0045] In parameter optimization, The modulation coefficient is linearly correlated with the neuronal excitability threshold of the subject. The gain coefficient is related to the magnetic susceptibility of the brain region. All online learning is performed under safety constraints. Candidate parameters undergo safety checks such as energy, peak value, and cumulative dose to prevent exceeding safety boundaries.

[0046] During stimulus execution and closed-loop feedback, the ultrasound transducer emits ultrasound waves with adjustable pulses, a static magnetic field is applied fixedly via a permanent magnet, and an electrode array continuously acquires EEG signals to achieve a closed-loop data flow. The hardware includes an electromagnetic shielding layer (μ-metal or permalloy, 1–3 mm thick) and twisted-pair shielded wires (shielding efficiency ≥60 dB), ensuring a signal SNR ≥20 dB and supporting local temperature rise monitoring, maximum energy, and session duration limits, complying with ICNIRP and relevant ultrasound safety standards. Dynamic adjustment strategies are implemented through real-time monitoring. value, and Compared with a set threshold, stimulation is terminated when the indicator reaches a stable condition for N consecutive time windows, and the system enters a low-power standby mode, while recording data for subsequent personalized optimization; if subsequent... If the value exceeds the threshold again, the system can quickly restore to the most recent valid parameter without reinitialization.

[0047] In summary, this embodiment achieves integrated control of the entire process from EEG signal acquisition, preprocessing, feature extraction to closed-loop stimulation and adaptive parameter optimization. Through high-resolution, multi-mode acquisition, dynamic threshold determination, and adaptive adjustment of safety constraints, it effectively solves the limitations of traditional transcranial stimulation methods in terms of real-time feedback, individualized adaptation, artifact interference, and safety control. It realizes precise, controllable, and personalized intervention of neural activity in the target brain region, providing a repeatable, safe, and efficient technical means for neuromodulation research and clinical applications.

[0048] It should be noted that the above description describes some embodiments of this application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in a different order than that shown in the above embodiments and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0049] Based on the same inventive concept, and corresponding to any of the above embodiments, the embodiments of this application also provide a closed-loop feedback system for transcranial magnetoacoustic stimulation guided by electroencephalography.

[0050] like Figure 5 As shown, the EEG-guided transcranial magnetic acoustic stimulation closed-loop feedback system includes: The signal acquisition module 11 is configured to acquire raw EEG signals from the target brain region and perform signal preprocessing; wherein, the signal preprocessing includes bandpass filtering, adaptive noise cancellation, and artifact removal; The feature extraction module 12 is configured to extract the activity features of the target brain region based on the preprocessed EEG signal, and determine the neural activity state of the target brain region based on the activity features of the target brain region; wherein, the power ratio of the target brain region is used as the core judgment index, and the phase difference and amplitude fluctuation range of the neural signals between the target brain region and the non-target brain region are calculated to dynamically evaluate the temporal synchronization and intensity changes of neural activity. The parameter generation and optimization module 13 is configured to generate and optimize the initial stimulus parameters based on the state determination result and a preset stimulus response model. The magnetoacoustic stimulation execution module 14 is configured to convert the generated initial stimulation parameters into actual stimulation signals, perform ultrasound and magnetic field stimulation on the target brain region by controlling the transcranial magnetoacoustic stimulation device, and provide closed-loop feedback based on the activity state of the target brain region.

[0051] For ease of description, the above apparatus is described in terms of its functions, divided into various modules. Of course, in implementing the embodiments of this application, the functions of each module can be implemented in one or more software and / or hardware.

[0052] The apparatus of the above embodiments is used to implement the corresponding method in any of the foregoing embodiments and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0053] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, embodiments of this application also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the methods described in any of the above embodiments.

[0054] Figure 6 This embodiment illustrates a more specific hardware structure of an electronic device, which may include a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, memory 1020, input / output interface 1030, and communication interface 1040 are interconnected internally via the bus 1050.

[0055] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0056] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program code is stored in the memory 1020 and is called and executed by the processor 1010.

[0057] The input / output interface 1030 is used to connect input / output modules to realize information input and output. The input / output modules can be configured as components in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touch screens, microphones, various sensors, etc., and output devices may include displays, speakers, vibrators, indicator lights, etc.

[0058] The communication interface 1040 is used to connect a communication module (not shown in the figure) to enable communication between this device and other devices. The communication module can communicate via wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0059] Bus 1050 includes a pathway for transmitting information between various components of the device, such as processor 1010, memory 1020, input / output interface 1030, and communication interface 1040.

[0060] It should be noted that although the above-described device only shows the processor 1010, memory 1020, input / output interface 1030, communication interface 1040, and bus 1050, in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the embodiments of this specification, and not necessarily all the components shown in the figures.

[0061] The electronic devices described above are used to implement the corresponding methods in any of the foregoing embodiments and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0062] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this application also provides a non-transitory computer-readable storage medium that stores computer instructions for causing the computer to perform the methods described in any of the above embodiments.

[0063] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0064] The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to perform the methods described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0065] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of this application (including the claims) is limited to these examples; within the framework of this application, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of the embodiments of this application as described above, which are not provided in the details for the sake of brevity.

[0066] Additionally, to simplify the description and discussion, and to avoid obscuring the embodiments of this application, the well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided drawings. Furthermore, the apparatus may be shown in block diagram form to avoid obscuring the embodiments of this application, and this also takes into account the fact that the details of the implementation of these block diagram apparatuses are highly dependent on the platform on which the embodiments of this application will be implemented (i.e., these details should be fully understood by those skilled in the art). While specific details (e.g., circuits) have been set forth to describe exemplary embodiments of this application, it will be apparent to those skilled in the art that the embodiments of this application can be implemented without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0067] Although this application has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0068] The embodiments of this application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the embodiments of this application should be included within the protection scope of this application.

Claims

1. A closed-loop feedback method for transcranial magnetoacoustic stimulation based on electroencephalography (EEG), characterized in that, include: Raw electroencephalogram (EEG) signals from the target brain region are acquired and preprocessed; wherein, the signal preprocessing includes bandpass filtering, adaptive noise cancellation, and artifact removal; Based on the preprocessed EEG signals, the activity characteristics of the target brain region are extracted, and the neural activity state of the target brain region is determined according to the activity characteristics of the target brain region. Among them, the power ratio of the target brain region is used as the core determination index. The phase difference and amplitude fluctuation range of the neural signals between the target brain region and the non-target brain region are calculated to dynamically evaluate the temporal synchronization and intensity changes of neural activity. Based on the state determination results, the initial stimulus parameters are generated and optimized in conjunction with the preset stimulus response model. The generated initial stimulation parameters are converted into actual stimulation signals. Ultrasonic and magnetic field stimulation is performed on the target brain region by controlling the transcranial magnetoacoustic stimulation device, and closed-loop feedback is performed based on the activity status of the target brain region.

2. The method according to claim 1, characterized in that: By setting a regular recording mode and a stimulus-synchronous high sampling mode, mode switching or parallel acquisition can be performed during the experiment; wherein, the stimulus-synchronous high sampling mode is used for modeling stimulus artifacts, generating templates, and performing post-processing.

3. The method according to claim 1, characterized in that: By performing time-frequency analysis on the preprocessed EEG signals, short-time Fourier transform or continuous wavelet transform is used to extract the power spectral density of the target brain region in a specific frequency band, and the power ratio of the target brain region is obtained based on the power spectral density. The phase difference and amplitude fluctuation range of the neural signal are obtained by using a sliding time window to determine whether there is abnormal synchronization across regions.

4. The method according to claim 1, characterized in that: The initial stimulation parameters include ultrasound pulse intensity, static magnetic field intensity, and stimulation frequency; wherein, the formula for the initial stimulation parameters is as follows: ； ； ； In the formula, , , These represent the ultrasonic pulse intensity, static magnetic field strength, and stimulation frequency, respectively. , , These are the reference values ​​for ultrasonic pulse intensity, static magnetic field intensity, and stimulation frequency. , , For adjustment coefficients, Indicates the range of amplitude fluctuations. Power spectral density in the frequency band, Indicates the phase difference of neural signals; By introducing an adaptive control mechanism to adjust the initial stimulus parameters online, the model can be dynamically corrected during closed-loop operation.

5. The method according to claim 1, characterized in that: The transcranial magnetoacoustic stimulation device includes an ultrasonic transducer and a permanent magnet. An electromagnetic shielding layer is provided between the ultrasonic transducer and the permanent magnet. The ultrasonic transducer emits ultrasonic waves in the form of pulses, and the static magnetic field is applied to the target brain region through the permanent magnet. Real-time EEG signals are acquired through an electrode array, with the electrodes using twisted-pair shielded cables.

6. The method according to claim 1, characterized in that: The power ratio of the target brain region, the phase difference of the neural signal, and the amplitude fluctuation range extracted in real time are compared with the preset dynamic adjustment threshold. In response to the determination of an abnormal state, the ultrasound intensity, static magnetic field intensity and stimulation frequency are dynamically adjusted based on the deviation of the current target brain region power ratio, neural signal phase difference and amplitude fluctuation range, in order to update the stimulation signal and continue to intervene until the termination condition is met. In response to the determination of a normal state, the stimulation is terminated and the system enters a low-power standby mode.

7. The method according to claim 6, characterized in that: Response to the continuous power ratio of the target brain region When all time windows are less than the preset threshold, and the absolute change rate of the fluctuation range of the neural signal phase difference and the amplitude fluctuation range both meet the stability threshold, the transcranial magnetic acoustic stimulation device automatically enters a low-power standby mode and records the EEG data of the intervention process.

8. A closed-loop feedback system for transcranial magnetoacoustic stimulation guided by electroencephalography (EEG), characterized in that, include: The signal acquisition module is configured to acquire raw electroencephalogram (EEG) signals from the target brain region and perform signal preprocessing; wherein, the signal preprocessing includes bandpass filtering, adaptive noise cancellation, and artifact removal; The feature extraction module is configured to extract activity features of the target brain region based on the preprocessed EEG signal, and determine the neural activity state of the target brain region based on the activity features of the target brain region; wherein, the power ratio of the target brain region is used as the core determination index, and the phase difference and amplitude fluctuation range of the neural signals between the target brain region and the non-target brain region are calculated to dynamically evaluate the temporal synchronization and intensity changes of neural activity. The parameter generation and optimization module is configured to generate and optimize initial stimulus parameters based on the state determination results and a preset stimulus response model. The magnetoacoustic stimulation execution module is configured to convert the generated initial stimulation parameters into actual stimulation signals, control the transcranial magnetoacoustic stimulation device to perform ultrasound and magnetic field stimulation on the target brain region, and provide closed-loop feedback based on the activity state of the target brain region.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the method as claimed in any one of claims 1-7.

10. A non-transitory computer-readable storage medium, characterized in that, in, The non-transitory computer-readable storage medium stores computer instructions for causing a computer to perform the method described in any one of claims 1-7.

Images