Intelligent closed-loop vagus nerve stimulation regulation and control method and system

Abstract

The invention relates to the technical field of medical equipment, in particular to an intelligent closed-loop vagus nerve stimulation regulation and control method and system. The method comprises the following specific steps: collecting an electroencephalogram signal and a stimulation signal during real-time stimulation, and obtaining an electroencephalogram signal prediction value of future N steps through an ambulatory electroencephalogram signal prediction model; calculating a depression biomarker estimated value for representing the brain activity state at the next moment according to the electroencephalogram signal predicted values of the next N steps; calculating an optimal stimulation signal at the next time according to a difference value between the depression biomarker target value and the depression biomarker estimated value; and updating the stimulation signal according to the optimal stimulation signal to stimulate the vagus nerve, and updating the stimulation signal. Meanwhile, the invention discloses a system for executing the method, the optimal stimulation parameter is calculated according to the difference between the predicted brain activity state at the next moment and the target state, so that the neural activity is controlled, and the adjusting efficiency and the control precision of the percutaneous vagus nerve stimulation system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical equipment, and in particular to an intelligent closed-loop vagus nerve stimulation control method and system. Background Art

[0002] Currently, the treatment of depression is mainly based on drug therapy, combined with psychotherapy, physical therapy and other treatments. Vagus nerve stimulation (VNS) regulates brain function by electrically stimulating the vagus nerve, and is currently used for neurostimulation treatment of severe refractory depression (TRD). This technology uses implantable or non-implantable devices to send electrical signals to the vagus nerve, which can regulate brain areas related to emotion processing (such as the amygdala and prefrontal cortex) and improve the mood and attention of patients with depression. Early clinical trials have shown effective evidence of the antidepressant efficacy and durability of VNS as a treatment for TRD. It was also found that VNS helps improve the quality of life and suicide rate of patients with unipolar TRD, as well as improve depression in patients with bipolar TRD.

[0003] Patent publication number CN117899356B discloses an adaptive auris vagus nerve stimulation device. An EEG acquisition module collects the patient's real-time EEG indicator signals; an ECG acquisition module collects the patient's real-time heart rate variability signals; and a data processing module analyzes the real-time EEG indicator signals and real-time heart rate variability signals within a preset time period based on a preset correspondence to obtain an adaptive stimulation current value. The auris vagus nerve stimulation module applies a target stimulation current to the patient's auris vagus nerve and adjusts the currently applied target stimulation current based on the adaptive stimulation current value at preset time intervals. This achieves adaptive regulation of the vagus nerve closed-loop stimulation current, improving the sensitivity and accuracy of adaptive regulation. However, existing technologies only regulate the current amplitude and do not consider the frequency. Current regulation is also performed based on the difference between the current EEG and ECG signals and the target EEG and ECG signals. However, during the actual stimulation process, the EEG signal is constantly changing, and adjusting the current based on the current EEG signal results in low accuracy. This makes it impossible to achieve optimal real-time stimulation. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent closed-loop vagus nerve stimulation control method and system to solve the above technical problems.

[0005] To achieve the above objectives, the present invention provides an intelligent closed-loop vagus nerve stimulation and control method, the specific steps of which are as follows:

[0006] Step S1: collecting resting-state EEG signals before stimulation, confirming the target brain activity state based on the resting-state EEG signals, and obtaining a target value of a depression biomarker based on the target brain activity state;

[0007] Step S2: collecting EEG signals and stimulation signals during real-time stimulation, preprocessing and frequency band decomposing the EEG signals collected during stimulation, extracting power features, and normalizing the power features;

[0008] Step S3: inputting the collected stimulation signal and the normalized power feature into the dynamic EEG signal prediction model to obtain the EEG signal prediction value for the next N steps;

[0009] Step S4: Calculating an estimated value of a depression biomarker for characterizing the brain activity state at the next moment based on the predicted value of the EEG signal N steps in the future;

[0010] Step S5: calculating the optimal stimulation signal for the next time period based on the difference between the target value of the depression biomarker and the estimated value of the depression biomarker, and updating the stimulation signal based on the optimal stimulation signal to stimulate the vagus nerve;

[0011] Step S6: Update the stimulation signal and repeat steps S2 to S5.

[0012] Preferably, in step S2, the preprocessing includes filtering and removing artifacts; extracting power features of different frequency bands of the EEG signal and calculating the average power in the overlapping time window, removing the mean and dividing by the standard deviation to obtain a standardized power feature.

[0013] Preferably, the training process of the dynamic EEG signal prediction model is as follows:

[0014] Acquire the data set: Stimulate the brain network with a parameter-randomized waveform with equal power across the entire frequency band. The stimulus wave contains amplitude and frequency data. Get the full-band brain network dynamic EEG signal Y t , all input stimuli U for each trial t ={u1,u2,...,u t} and output signal Y t ={y1,y2,...,y t} is composed of a dataset {u t ,y t};

[0015] Divide the dataset: The above dataset is divided into training set and test set using four-fold cross validation;

[0016] Construct a dynamic EEG signal prediction model, which is as follows:

[0017]

[0018] Among them, y t is the EEG signal of the t-th time segment, xt To use artificial neural network from y t The latent state extracted from is the potential representation set of brain functional network states, x t+1 To obtain the value of y through artificial neural network t+1 The potential state extracted from ; A, B and C are the hidden state feature matrix, input matrix and output matrix respectively;

[0019] Training and testing: The constructed dynamic EEG signal prediction model is trained using the training set until convergence or the number of iterations is reached, and then tested using the test set until the accuracy evaluation index of the dynamic EEG signal prediction model reaches the set value.

[0020] Preferably, an estimated value of a depression biomarker is calculated based on the predicted value of the electroencephalogram signal in the next N steps, and the estimated value of the depression biomarker includes individual alpha band peak frequency and relative gamma band power.

[0021] Preferably, step S5 is specifically as follows:

[0022] Construct a cost function based on the target brain activity state. The cost function j is as follows:

[0023]

[0024] Among them, s t is the estimated value of depression biomarker at time t, For s t The transpose of s t+N is the estimated value of depression biomarker predicted at step N, For s t+N The transpose of , Q is the state weight matrix, which adjusts the importance of different state variables, R is the control input weight matrix, which adjusts the degree of penalty for the control input amplitude, and F is the terminal state weight matrix, which is used to ensure the stability of the system at the end of the prediction period;

[0025] Solve minJ under the constraints of the safe stimulus parameter range, where min is the minimum value, and obtain the control sequence u in the finite time domain. t+1 ~u t+N-1 , take u t+1 As the optimal parameters of the input stimulus for the next time period

[0026] Preferably, the optimal parameters Merged into U t Get U t+1 , the EEG signal y measured after stimulation t+1 Merged into Y t Get Y t+1, use the new data set to refit the dynamic EEG signal prediction model and update the EEG signal prediction value for the next N steps.

[0027] A system for implementing an intelligent closed-loop vagus nerve stimulation control method, comprising:

[0028] An electrical stimulation module, configured to output a stimulation current according to the stimulation signal, wherein stimulation parameters of the stimulation current include current intensity and oscillation frequency;

[0029] EEG signal acquisition modules, wherein several EEG signal acquisition modules are integrated on the wearable device, and several EEG signal acquisition modules correspond to closed-loop vagus nerves at different locations, and several EEG signal acquisition modules are connected to a multiplexer or several analog-to-digital converters with amplification functions;

[0030] The control and calculation module is used for signal processing, building and training dynamic EEG signal prediction models, and calculating the optimal stimulation signal;

[0031] A signal transmission module is used to realize communication between the EEG signal acquisition module and the control and calculation module, and between the electrical stimulation module and the control and calculation module;

[0032] The coordination module, the electrical stimulation module, the electroencephalogram signal acquisition module, the control and calculation module, and the signal transmission module are all electrically connected to the coordination module to achieve coordinated control and event recording of the electrical stimulation module, the electroencephalogram signal acquisition module, the control and calculation module, and the signal transmission module.

[0033] Therefore, the present invention adopts the above-mentioned intelligent closed-loop vagus nerve stimulation control method and system, which has the following beneficial effects:

[0034] (1) The collected stimulation signal is input into the dynamic EEG signal prediction model to obtain the EEG signal prediction value for the next N steps. The depression biomarker estimation value used to characterize the brain activity state at the next moment is calculated based on the EEG signal prediction value for the next N steps. The optimal stimulation signal at the next moment is calculated based on the difference between the depression biomarker target value and the depression biomarker estimation value, thereby realizing dynamic adjustment of the stimulation signal with high accuracy.

[0035] (2) The control parameters in the stimulation signal include amplitude and frequency to achieve more precise functional regulation.

[0036] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is a flow chart of an intelligent closed-loop vagus nerve stimulation and control method of the present invention;

[0038] Figure 2This is a system block diagram of the present invention. DETAILED DESCRIPTION

[0039] In the description of the present invention, it should be noted that the terms "upper", "lower", "inside", "outside" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, or the orientations or positional relationships in which the inventive product is usually placed when in use. These are only for the convenience of describing the present invention and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, they cannot be understood as limitations on the present invention. In the description of the present invention, it should also be noted that, unless otherwise expressly specified and limited, the terms "setting", "installation" and "connection" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be a communication between the internal parts of two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0040] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0041] like Figure 1 As shown, an intelligent closed-loop vagus nerve stimulation control method, the specific steps are as follows:

[0042] Step S1: Collect resting-state EEG signals before stimulation, confirm the target brain activity state based on the resting-state EEG signals, and obtain the target value of the depression biomarker based on the target brain activity state.

[0043] Step S2: During real-time stimulation, EEG signals and stimulation signals are collected. The EEG signals collected during stimulation are preprocessed and frequency-band decomposition is performed. Preprocessing includes filtering and artifact removal. Filtering filters the EEG signals to a frequency band of 1-45 Hz. Stimulation artifacts are identified and removed to obtain pure EEG signal data.

[0044] We extracted power features from different frequency bands of the EEG signal and calculated the average power within overlapping time windows (1-8 Hz, 8-12 Hz, 12-30 Hz, and 30-45 Hz). The overlapping time windows were 10 seconds long with a 1-second time step. After removing the mean and dividing by the standard deviation, we obtained standardized power features suitable for subsequent analysis.

[0045] Step S3: Input the collected stimulation signal and the normalized power feature into the dynamic EEG signal prediction model to obtain the EEG signal prediction value for the next N steps.

[0046] The training process of the dynamic EEG signal prediction model is as follows:

[0047] Acquiring the data set: Stimulating the brain network with a parameter-randomized waveform with equal power across the entire frequency band. In actual implementation, a continuous stimulation method is used. Every 1 second, a set of stimulation waveforms determined by amplitude and frequency parameters are randomly selected for vagus nerve stimulation. The stimulation wave contains amplitude and frequency data. The amplitude and frequency combinations of the electrical stimulation waveforms include:

[0048] {(0mA,0Hz),(150mA,5Hz),(150mA,20Hz),(280mA,5Hz),

[0049] (280mA, 20Hz)}, one of them is randomly selected each time stimulation occurs, and the probability of each combination being selected is A total of 10 trials with exactly the same stimulation scheme were conducted, each trial was stimulated 180 times, each stimulation lasted 1 second, and the 10 trials lasted a total of 30 minutes. The full-band brain network dynamic EEG signal Y was obtained. t , all input stimuli U for each trial t ={u1,u2,...,u t} and output signal Y t ={y1,y2,...,y t} is composed of a dataset {u t ,y t};

[0050] Dataset division: The above dataset is divided into training set and test set using four-fold cross-validation; for each cross-validation, three-quarters of each test signal is divided into training set, concatenated head to tail, and the remaining signal of each test is used as test set.

[0051] Construct a dynamic EEG signal prediction model, which is as follows:

[0052]

[0053] Among them, y t is the EEG signal of the t-th time segment, x t To obtain the value of y through artificial neural network t The latent state extracted from is the potential representation set of brain functional network states, x t+1 To use artificial neural network from y t+1 The latent state extracted from the latent space of the brain network is obtained; A, B and C are the hidden state feature matrix, input matrix and output matrix respectively; the dynamic pattern decomposition method is used to estimate the parameters of the hidden state feature matrix A, input matrix B and output matrix C of the latent space of the brain network to characterize the overall dynamic state of the brain network.

[0054] Training and testing: The constructed dynamic EEG signal prediction model is trained on the training set until convergence or the number of iterations is reached, and then tested on the test set until the accuracy evaluation index of the dynamic EEG signal prediction model reaches the set value. The linear correlation coefficient (LCC) between the predicted brain network dynamics, i.e., the EEG signal power characteristics, and the actual signal power characteristics is calculated on the test set as a quantitative indicator of model accuracy. The linear correlation coefficient calculation formula is as follows:

[0055]

[0056] in, is the EEG signal power eigenvalue predicted by the model, is the actual EEG signal power eigenvalue, Cov() is the covariance function, and Var() is the variance function.

[0057] Step S4: Calculate an estimated depression biomarker value, representing the brain activity state at the next moment, based on the predicted EEG signal values ​​N steps in the future. This depression biomarker estimate includes individual alpha-band peak frequency, relative gamma-band power, and other parameters. The baseline is defined as the biomarker estimate of clinically normal brain network activity.

[0058] Step S5: Calculate the optimal stimulation signal for the next time according to the difference between the target value of the depression biomarker and the estimated value of the depression biomarker.

[0059] Construct a cost function based on the target brain activity state. The cost function J is as follows:

[0060]

[0061] Among them, s t is the estimated value of depression biomarker at time t, For s t The transpose of s t+N is the estimated value of depression biomarker predicted at step N, For s t+N Q is the state weight matrix, which adjusts the importance of different state variables. R is the control input weight matrix, which adjusts the degree of penalty for the control input amplitude. F is the terminal state weight matrix, which is used to ensure the stability of the system at the end of the prediction period.

[0062] Solve minJ under the constraints of the safe stimulus parameter range, where min is the minimum value, and obtain the control sequence u in the finite time domain. t+1 ~u t+N-1 , take u t+1 As the optimal parameters of the input stimulus for the next time period The stimulation signal is updated according to the optimal stimulation signal to stimulate the vagus nerve.

[0063] Step S6: Update the stimulation signal and repeat steps S2 to S5.

[0064] The optimal parameters Merged into U t Get U t+1 , the EEG signal y measured after stimulation t+1 Merged into Y t Get Y t+1 , use the new data set to refit the dynamic EEG signal prediction model and update the EEG signal prediction value for the next N steps.

[0065] A system for executing the above-mentioned intelligent closed-loop vagus nerve stimulation control method, such as Figure 2 Shown, including:

[0066] The electrical stimulation module is used to output stimulation current according to the stimulation signal. It can apply a variety of waveforms including direct current (tDCS), oscillating current of specified frequency (tACS), and pulsed or modulated current. The stimulation parameters of the stimulation current include current intensity and oscillation frequency. The electrical stimulation module has a parameter adjustment function, which can adjust the stimulation current according to the optimal parameters. It usually includes a controllable current source or voltage source. The design uses a bipolar constant current driver to generate current between any two electrodes, and the channel is switched by an analog switch network. The electrical stimulation module has an electrode impedance detection function, which injects a small test current or uses the existing stimulation current to measure the impedance. If poor electrode contact is detected (impedance is too high), the system will stop stimulation to prevent voltage overload.

[0067] The EEG signal acquisition module consists of a set of electrodes and related front-end electronic components. The electrodes are dry contact sensors made of conductive foam or semi-dry sponge materials to improve poor contact, increase convenience, enhance user comfort, and speed up device installation. Several EEG signal acquisition modules are integrated into the wearable device, and several EEG signal acquisition modules correspond to closed-loop vagus nerves at different locations. This embodiment uses 8 channels of EEG signals, including Fp1 (left frontal pole), Fp2 (right frontal pole), F7 (left anterior temporal), F8 (right anterior temporal), T3 (left parietal lobe), T4 (right parietal lobe), P3 (left middle temporal), and P4 (right middle temporal). It can be expanded according to actual needs. Users can connect additional sensor modules at predetermined locations on the head-mounted device. The device will automatically recognize the new electrodes and incorporate them into the EEG signal acquisition module array, thereby increasing the number of channels or covering additional areas. The EEG signals of all active electrodes are output through a multiplexer or an analog-to-digital converter with amplification function.

[0068] The control and calculation module includes a model prediction controller and a display and operation external device. The display and operation external device allows clinicians or users to view the system status and the settings of related parameters. The model prediction controller is used for signal processing, building and training dynamic electroencephalogram signal prediction models, calculating optimal stimulation signals, and transmitting the calculated optimal stimulation parameters to the electrical stimulation module through the signal transmission module to execute the adjusted stimulation.

[0069] The control computing module can also connect to cloud services to enhance processing power and data storage, and can also run offline without relying on external services. This makes the entire closed-loop system portable and wearable, which has significant advantages over solutions that rely on laboratory equipment or magnetic resonance imaging.

[0070] The signal transmission module facilitates communication between the EEG signal acquisition module and the control and computation module, as well as between the electrical stimulation module and the control and computation module. Wireless transmission utilizes a low-power Bluetooth module, Wi-Fi module, 4G module, or 5G module. Alternatively, direct data transmission can be achieved using a wired interface, such as a USB port. The transmission process includes error detection and encryption to ensure the privacy of EEG data. The time from EEG acquisition to stimulation command can be controlled within 100 milliseconds.

[0071] The coordination module includes a microcontroller or an embedded processor. The electrical stimulation module, the electroencephalogram signal acquisition module, the control and calculation module, and the signal transmission module are all electrically connected to the coordination module to achieve coordinated control and event recording of the electrical stimulation module, the electroencephalogram signal acquisition module, the control and calculation module, and the signal transmission module.

[0072] For example, exceeding a certain current limit (eg, 2 mA maximum) or a certain duration is not allowed unless the safety authentication mode is explicitly enabled.

[0073] For transcranial direct current stimulation, gradual increases or decreases can prevent user discomfort, and the current can be automatically and smoothly increased or decreased to avoid sudden changes.

[0074] If communication with the control and computation modules is interrupted, the coordination module can immediately terminate any ongoing stimulation or revert to the default program.

[0075] The coordination module can also perform basic signal processing—for example, it might calculate coarse EEG metrics, such as whether the EEG signal is within a normal range, to serve as a backup trigger mechanism to ensure that the user's brain activity remains within a safe range during closed-loop operation.

[0076] The coordination module contains memory to record events (when a stimulus was applied, parameters were set, etc.) for later analysis or to comply with regulatory requirements.

[0077] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the same. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solutions of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. An intelligent closed-loop vagus nerve stimulation control method, characterized in that: The specific steps are as follows: Step S1: collecting resting-state EEG signals before stimulation, confirming the target brain activity state based on the resting-state EEG signals, and obtaining a target value of a depression biomarker based on the target brain activity state; Step S2: collecting EEG signals and stimulation signals during real-time stimulation, preprocessing and frequency band decomposing the EEG signals collected during stimulation, extracting power features, and normalizing the power features; Step S3: inputting the collected stimulation signal and the normalized power feature into the dynamic EEG signal prediction model to obtain the EEG signal prediction value for the next N steps; Step S4: Calculating an estimated value of a depression biomarker for characterizing the brain activity state at the next moment based on the predicted value of the EEG signal N steps in the future; Step S5: calculating the optimal stimulation signal for the next time period based on the difference between the target value of the depression biomarker and the estimated value of the depression biomarker, and updating the stimulation signal based on the optimal stimulation signal to stimulate the vagus nerve; Step S6: Update the stimulation signal and repeat steps S2 to S5.

2. The intelligent closed-loop vagus nerve stimulation control method according to claim 1, characterized in that: In step S2, preprocessing includes filtering and removing artifacts; extracting power features of different frequency bands of the EEG signal and calculating the average power in the overlapping time window, removing the mean and dividing by the standard deviation to obtain the standardized power feature.

3. The intelligent closed-loop vagus nerve stimulation control method according to claim 2, characterized in that: The training process of the dynamic EEG signal prediction model is as follows: Acquire the data set: Stimulate the brain network with a parameter-randomized waveform with equal power across the entire frequency band. The stimulus wave contains amplitude and frequency data. Get the single-channel full-band brain network dynamic EEG signal Y t , all input stimuli U for each trial t ={u1,u2,...,u t } and single channel output signal Y t ={y1,y2,...,y t } is composed of a dataset {u t ,y t }; Divide the dataset: The above dataset is divided into training set and test set using four-fold cross validation; Construct a dynamic EEG signal prediction model, which is as follows: Among them, y t is the EEG signal of the t-th time segment, x t To use artificial neural network from y t The latent state extracted from is the potential representation set of brain functional network states, x t+1 To use artificial neural network from y t+1 The potential state extracted from ; A, B and C are the hidden state feature matrix, input matrix and output matrix respectively; Training and testing: The constructed dynamic EEG signal prediction model is trained using the training set until convergence or the number of iterations is reached, and then tested using the test set until the accuracy evaluation index of the dynamic EEG signal prediction model reaches the set value.

4. The intelligent closed-loop vagus nerve stimulation control method according to claim 3, characterized in that: The estimated value of depression biomarkers is calculated based on the predicted value of the EEG signal in the next N steps. The estimated value of depression biomarkers includes the individual α-band peak frequency and relative γ-band power.

5. The intelligent closed-loop vagus nerve stimulation control method according to claim 4, characterized in that: Step S5 is specifically as follows: Construct a cost function based on the target brain activity state. The cost function J is as follows: Among them, s t is the estimated value of depression biomarker at time t, For s t The transpose of s t+N is the estimated value of depression biomarker predicted at step N, For s t+N The transpose of , Q is the state weight matrix, which adjusts the importance of different state variables, R is the control input weight matrix, which adjusts the degree of penalty for the control input amplitude, and F is the terminal state weight matrix, which is used to ensure the stability of the system at the end of the prediction period; Solve minJ under the constraints of the safe stimulus parameter range, where min is the minimum value, and obtain the control sequence u in the finite time domain. t+1 ~u t+N-1 , take u t+1 As the optimal parameters of the input stimulus for the next time period 6. The intelligent closed-loop vagus nerve stimulation control method according to claim 5, characterized in that: The optimal parameters Merged into U t Get U t+1 , the EEG signal y measured after stimulation t+1 Merged into Y t Get Y t+1 , use the new data set to refit the dynamic EEG signal prediction model and update the EEG signal prediction value for the next N steps.

7. A system for executing the intelligent closed-loop vagus nerve stimulation control method according to any one of claims 1 to 6, characterized in that: include: An electrical stimulation module, configured to output a stimulation current according to the stimulation signal, wherein stimulation parameters of the stimulation current include current intensity and oscillation frequency; EEG signal acquisition modules, wherein several EEG signal acquisition modules are integrated on the wearable device, and several EEG signal acquisition modules correspond to closed-loop vagus nerves at different locations, and several EEG signal acquisition modules are connected to a multiplexer or several analog-to-digital converters with amplification functions; The control and calculation module is used for signal processing, building and training dynamic EEG signal prediction models, and calculating the optimal stimulation signal; A signal transmission module is used to realize communication between the EEG signal acquisition module and the control and calculation module, and between the electrical stimulation module and the control and calculation module; The coordination module, the electrical stimulation module, the electroencephalogram signal acquisition module, the control and calculation module, and the signal transmission module are all electrically connected to the coordination module to achieve coordinated control and event recording of the electrical stimulation module, the electroencephalogram signal acquisition module, the control and calculation module, and the signal transmission module.

Images