Transcranial electrical stimulation physiotherapy system based on EEG-fNIRS feedback

Abstract

The invention discloses a transcranial electrical stimulation physiotherapy system based on EEG-fNIRS feedback. The transcranial electrical stimulation physiotherapy system comprises an EEG-fNIRS signal acquisition and processing module, a relaxation degree evaluation module, a transcranial electrical stimulation control module and a transcranial electrical stimulation execution module. The transcranial electrical stimulation physiotherapy system provided by the invention can accurately extract spatial and temporal characteristics related to relaxation degree, constructs a personalized brain function network model, accurately locates stimulation sites and neural activity modes based on the characteristic that EEG and fNIRS data complement each other in terms of advantages in terms of time resolution and spatial resolution, and improves the rehabilitation effect of the brain function network model on the basis of the characteristic that the advantages of the EEG and fNIRS data complement each other in terms of time resolution and spatial resolution. The method realizes personalized dynamic adjustment of the relaxation degree, and is suitable for multiple fields of education, medical treatment, rehabilitation and the like.

Description

Technical Field

[0001] The present invention relates to the field of neuromodulation, and more specifically, to a transcranial electrical stimulation physiotherapy system based on EEG-fNIRS feedback. Background Art

[0002] Meditation is an ancient mental training method that helps people achieve a state of relaxation and tranquility by focusing on the present moment and excluding distracting thoughts. Research has shown that meditation helps reduce stress, improve concentration, enhance mood, and increase self-awareness. However, beginners often have difficulty concentrating during meditation and are easily distracted by external factors or internal thoughts. In addition, finding a suitable meditation method and maintaining a long-term practice habit are also major challenges. otACS (transcranial alternating current stimulation) uses alternating current, the frequency of which can be adjusted to specific brain wave bands, such as Alpha or Theta waves. These bands are closely related to the relaxation and meditation states. By stimulating specific brain wave bands related to relaxation, otACS can help users reach the meditation state faster and enhance brain activity during meditation. At the same time, by applying a low-intensity current to the cerebral cortex, otACS can regulate the excitability of neurons, thereby achieving effects such as alleviating depressive symptoms, enhancing cognitive function, reducing migraines, assisting in rehabilitation treatment, improving sleep quality, and promoting attention regulation. Through precise electrical stimulation regulation, otACS can optimize the neural activity of specific brain regions, thereby improving the targeting and effectiveness of treatment.

[0003] Currently, due to individual differences and regulation cycle differences, the fixed-mode otACS (transcranial alternating current stimulation) cannot accurately meet the needs of users at different intervention periods. The causal closed-loop of stimulation-monitoring-feedback adjustment helps to adjust the stimulation parameters in real time and formulate a stimulation plan according to the brain activity pattern of each patient. For example, a method and system for intelligent neuromodulation based on attention state disclosed by Liu Yulong in the invention patent CN118925061A. An EEG signal is evaluated using a relaxation degree evaluation model to judge the user's current attention state, and neuromodulation is carried out according to the evaluation result. However, there is still much room for improvement in terms of regulation parameters and regulation sites in this method. Existing neuromodulation programs are usually based on standardized parameter configurations and lack a dynamic adjustment mechanism for individual differences, resulting in differences in regulation effects among different individuals and different intervention stages. Summary of the Invention

[0004] To solve the above technical problems, the purpose of the present invention is to provide a technical solution for synchronously performing personalized neuromodulation on subjects by integrating EEG and fNIRS technologies, so as to improve the accuracy and intervention effect of neuromodulation.

[0005] The transcranial electrical stimulation physiotherapy system based on EEG-fNIRS feedback includes: an EEG-fNIRS signal acquisition and processing module, which is used to synchronously acquire and record EEG and fNIRS data of a subject in a relaxed state, and train a relaxation degree evaluation model based on the EEG and fNIRS data; a relaxation degree evaluation module, which is used to synchronously acquire and record real-time EEG and fNIRS data of the subject, evaluate the relaxation degree of the subject through the relaxation degree evaluation model, and judge the relaxation degree state of the subject according to the result of the relaxation degree evaluation; a transcranial electrical stimulation control module, based on the relaxation degree state of the subject, determines the personalized intervention parameters of otACS, and a transcranial electrical stimulation execution module, which receives the personalized intervention parameters of otACS determined by the electrical stimulation control module, and performs personalized neural regulation on the subject through transcranial alternating current stimulation.

[0006] Preferably, the EEG-fNIRS signal acquisition and processing module includes an EEG-fNIRS signal acquisition module and an EEG-fNIRS signal processing module. The EEG signal acquisition module and the fNIRS signal acquisition module in the EEG-fNIRS signal acquisition module are set to have spatial identity; the collected EEG and fNIRS data are preprocessed such as re-referencing, filtering, and artifact removal, and then features related to the relaxation degree are extracted, and a relaxation degree evaluation model is trained based on a spatio-temporal feature convolutional network.

[0007] Preferably, the transcranial electrical stimulation control module determines the personalized intervention parameters of otACS, including the stimulation frequency, stimulation phase, stimulation intensity, and stimulation site of otACS.

[0008] Preferably, the transcranial electrical stimulation control module constructs a brain functional network model of the subject in a relaxed state based on causal analysis, extracts and records the network betweenness centrality of the subject in the relaxed state; combines the real-time EEG and fNIRS data of the subject, constructs a brain functional network model of the subject based on causal analysis, extracts the network betweenness centrality of the subject in real time, compares the network betweenness centrality determined in real time with the network betweenness centrality of the subject in the relaxed state, and selects the node with the largest difference in betweenness centrality as the priority stimulation site for intervention.

[0009] Preferably, the transcranial electrical stimulation control module extracts the resonance frequencies of individual Alpha oscillation and Theta oscillation by spectrum analysis as the stimulation frequency of otACS.

[0010] Preferably, the transcranial electrical stimulation control module extracts the resonance frequencies of the individual Alpha oscillation and Theta oscillation through the following steps:

[0011] Calculate the power spectral curves of Alpha oscillation and Theta oscillation for each data segment using the Fast Fourier Transform. Due to the differences between brain regions, the calculation is divided according to EEG leads;

[0012] Overlay the spectral energy values of the individual Alpha and Theta frequency bands at the data segment level respectively;

[0013] Find the maximum value of the average power spectral curve at the corresponding data segment level as the resonance frequencies of Alpha oscillation and Theta oscillation.

[0014] Preferably, the transcranial electrical stimulation control module extracts the instantaneous phases of the individual Alpha oscillation and Theta oscillation using phase estimation as the stimulation phases of otACS.

[0015] Preferably, the transcranial electrical stimulation control module extracts the otACS stimulation phases through the following steps: Use the spectral analysis method to obtain the power spectral density of the Alpha and Theta frequency bands in the individual EEG signal, and estimate the phase change of the oscillation signal through the autoregressive model; Use the estimated phase information to determine the instantaneous phases of Alpha and Theta oscillations at each moment.

[0016] Compared with the prior art, the present invention can at least achieve the following beneficial effects:

[0017] The transcranial electrical stimulation physiotherapy system based on EEG-fNIRS feedback provided by the present invention combines EEG and fNIRS technologies, provides the ability to monitor brain activities with high spatio-temporal resolution, can accurately extract spatio-temporal features related to relaxation, and construct a personalized brain function network model to realize the dynamic optimization of transcranial electrical stimulation parameters. Through the closed-loop feedback control mechanism, the stimulation intensity, frequency and phase are adjusted in real time to ensure the best intervention effect; This method has the advantages of non-invasiveness, wide applicability and data accumulation ability, and can significantly improve the adaptability and effectiveness of neuromodulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic structural composition diagram of a transcranial electrical stimulation physiotherapy system based on EEG-fNIRS feedback provided by an embodiment of the present invention.

[0019] Figure 2 It is the feature network framework of the spatio-temporal feature convolutional network provided by an embodiment of the present invention.

[0020] Figure 3 It is a flowchart for extracting the resonance frequencies of Alpha oscillation and Theta oscillation provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0022] In order to enable those skilled in the art to better understand the solutions of the present invention, the present invention will be further described in detail below in conjunction with the drawings and specific embodiments.

[0023] The present invention aims to provide a transcranial electrical stimulation physiotherapy system based on EEG-fNIRS feedback to better evaluate the relaxation state of the subject. Through transcranial alternating current stimulation, personalized neuroregulation is performed on the subject to make the subject enter a relaxed mental state, so as to achieve effects such as relieving the symptoms of depression in the subject, improving cognitive function, reducing migraine, assisting in rehabilitation treatment, improving sleep quality, and promoting attention regulation.

[0024] Reference Figure 1 , Figure 1 is a schematic structural composition diagram of a transcranial electrical stimulation physiotherapy system provided by an embodiment of the present invention.

[0025] A transcranial electrical stimulation physiotherapy system based on EEG-fNIRS feedback provided by the present invention includes: an EEG-fNIRS signal acquisition and processing module, a relaxation degree evaluation module, a transcranial electrical stimulation control module, and a transcranial electrical stimulation execution module.

[0026] The EEG-fNIRS signal acquisition and processing module includes an EEG-fNIRS signal acquisition module and an EEG-fNIRS signal processing module, and is used to synchronously acquire and record EEG and fNIRS data of the subject in a relaxed state, and train a relaxation degree evaluation model based on the EEG and fNIRS data.

[0027] Further, the EEG-fNIRS signal acquisition and processing module includes an EEG-fNIRS signal acquisition module and an EEG-fNIRS signal processing module, and the EEG signal acquisition module and the fNIRS signal acquisition module in the EEG-fNIRS signal acquisition module are spatially identical.

[0028] As one of the embodiments of the present invention, the EEG-fNIRS signal acquisition module adopts a multi-channel EEG-fNIRS synchronous acquisition device. The multi-channel EEG-fNIRS device should be equipped with EEG electrodes and fNIRS probes at the same time. For the same channel site, the EEG electrode patch and the fNIRS probe overlap, and can synchronously acquire the EEG and fNIRS signals at the same site in terms of time and space. The EEG-fNIRS synchronous acquisition device should overlap as much as possible in spatial distribution, and the sampling rate of EEG data should be at least 1000 Hz to avoid distortion of the original data caused by subsequent filtering operations. The EEG-fNIRS synchronous acquisition device has channels covering the parietal lobe, temporal lobe, and prefrontal lobe, and monitors brain activities related to relaxation such as spatial perception, memory, learning, and emotion. The EEG and fNIRS signal transmission circuits are isolated to avoid crosstalk between digital signals and analog signals. The fNIRS signal uses a pulse cycle mode to capture the near-infrared light intensity, with higher numerical accuracy. As one of the embodiments of the present invention, by artificially guiding the subject to perform nerve activation and meditation relaxation tasks, the subject is brought into a soothing and relaxing state, so as to collect the EEG and fNIRS data of the subject and train a relaxation degree evaluation model. It can be understood that the nerve activation task aims to stimulate the neuronal activity of the brain through specific cognitive or motor activities, so as to generate detectable electrophysiological signals and near-infrared spectral signals. Specific nerve activation tasks can include: cognitive tasks, such as mental arithmetic tasks, language understanding tasks, memory recall tasks, etc. These tasks require the subject to perform high-intensity cognitive processing to activate the neuronal activity in the relevant areas of the brain. The meditation relaxation task aims to help the subject achieve a state of physical and mental relaxation through meditation practice, so as to observe the electrophysiological and near-infrared spectral signal characteristics of the brain in a relaxed state.

[0029] As one of the embodiments of the present invention, by artificially guiding the subject to perform nerve activation and meditation relaxation tasks, the subject is brought into a soothing and relaxing state. The nerve activation task specifically refers to a mental arithmetic task. Under the guidance of the researcher, the wearer performs continuous addition (or subtraction) tasks. At the beginning of the task, the wearer obtains a four-digit initial number, and the task prompt requires adding (or subtracting) a single-digit number. During the task process, the wearer needs to continuously perform calculations, avoid external interference, and ensure that the brain is effectively activated in relatively simplified cognitive activities. After the neuronal activity of the subject is activated, the researcher provides meditation guidance to help the subject perform meditation practices such as deep breathing, relaxing various parts of the body, and clearing the mind, so that the subject enters a relaxed mental and physiological state, reducing the cognitive burden on the brain, and thus providing stable reference data for subsequent signal analysis.

[0030] The EEG-fNIRS signal processing module is used to perform preprocessing operations such as rereferencing, filtering, and artifact removal on the collected EEG and fNIRS data, extract relaxation-related features therefrom, and then train a relaxation evaluation model based on a spatio-temporal feature convolutional network.

[0031] It can be understood that the preprocessing operations are to perform preprocessing operations on the collected EEG signals and fNIRS signals respectively, and at the same time perform time alignment on the signals to ensure the synchronization of EEG and fNIRS data. The processing of EEG signals includes rereferencing, filtering, and artifact removal of EEG signals to eliminate noise interference such as electrooculogram and electromyogram. The processing of fNIRS signals: The collected fNIRS signals are used to extract the concentration changes of oxyhemoglobin (HbO) and deoxyhemoglobin (HbR) through optical density conversion and then perform normalization processing.

[0032] As one implementation manner of the present invention, refer to Figure 2 , Figure 2 , which is the feature network framework of the spatio-temporal feature convolutional network provided by the embodiments of the present invention. Specifically, for the spatio-temporal feature convolutional network, first, one-dimensional convolutional kernels are used to adjust the EEG and fNIRS data to the same dimension, and then the spatio-temporal feature convolutional network is imported to extract the EEG-fNIRS signal features. The spatio-temporal feature convolutional network includes a temporal convolutional kernel and a spatial convolutional kernel, which respectively extract features in the temporal dimension and the spatial dimension. As one embodiment of the present invention, specifically, the mathematical expression of the temporal convolutional kernel is as follows:

[0033] X' = Conv1d(X, W t ) + b t

[0034] where: X is the input EEG data, with a shape of (N, C, T), N is the number of samples, C is the number of channels (i.e., the number of electrodes), and T is the time step; W t is the temporal convolutional kernel, with a shape of (C out , 1, 1), where C out is the number of output channels; b t is the bias term, with a shape of (C out , 1, 1). X′ is the output after convolution, with a shape of (N, C out , T).

[0035] The spatial convolutional kernel performs convolution in the channel dimension and focuses on the features between channels. The mathematical expression of the spatial convolutional kernel is as follows:

[0036] Y = Conv2d(X, W s ) + b s

[0037] where: Ws is a spatial convolution kernel with a shape of (T out , C out , 1, 1), where T out is the time step and C out is the number of output channels; b s is the bias term with a shape of (T out , C out ). Y is the output after convolution with a shape of (N, T out , C out ).

[0038] The last layer is a fully connected layer, and the output dimension is the same as that of the labeled fNIRS data. The mean squared error is selected as the loss function, and the calculation is transformed into a regression problem.

[0039] Based on the test of the feature extraction network, as a preferred embodiment, the shape of the time convolution kernel is (1, 64), the number dimension is 8, the shape of the spatial convolution kernel is (C, 1), and the number is 4.

[0040] The relaxation degree evaluation module, specifically, when the subject starts physical therapy, the real-time EEG and fNIRS data of the subject collected by the EEG-fNIRS signal acquisition module are synchronously imported into the relaxation degree evaluation model for evaluation to determine whether the relaxation degree state of the subject is lower than the expected standard. If the evaluation result is lower than the expected standard, the transcranial electrical stimulation control module determines the otACS personalized intervention parameters based on the relaxation degree state of the subject;

[0041] As one implementation manner of the present invention, before the subject undergoes physical therapy, a corresponding relaxation degree threshold is set as the expected evaluation standard. Specifically, in one embodiment of the present invention, the relaxation degree evaluation model can directly generate the probability value that the current state of the subject approaches the human body's soothing and relaxing state based on the collected real-time EEG and fNIRS data of the subject, and the probability value shows an exponential law. In one embodiment of the present invention, the probability value is evaluated from 0 to 100 points, and the fluctuation of the probability value between 50 and 60 points is relatively large. Therefore, in one embodiment of the present invention, the relaxation degree threshold is set to 60 points. When the relaxation degree evaluation module evaluates that the relaxation degree of the subject is lower than 60 points, it feeds back to the transcranial electrical stimulation control module, and the transcranial electrical stimulation control module determines the otACS personalized intervention parameters based on the relaxation degree state of the subject.

[0042] The transcranial electrical stimulation control module determines the stimulation frequency, stimulation phase, stimulation intensity, and stimulation site of otACS personalized intervention based on the relaxation state of the subject. Further, as one embodiment of the present invention, the otACS personalized intervention parameters are determined through the following steps:

[0043] Step 1: Combine the real-time EEG and fNIRS data of the subject, construct a brain functional network model of the subject based on causal analysis, extract the betweenness centrality of the network, compare the real-time determined network betweenness centrality with the network betweenness centrality of the subject in a soothing and relaxed state, and select the node with the largest difference in betweenness centrality as the priority stimulation site for intervention.

[0044] It can be understood that the brain functional network model is constructed based on a causal matrix, where nodes represent electrodes or voxels, and edges represent the strength of the causal relationship between nodes. Through causal analysis, the causal relationship of neural activities between different regions of the brain can be determined. As one embodiment of the present invention, the Granger causal analysis method is adopted. Comparing the real-time determined network betweenness centrality with the network betweenness centrality of the subject in a soothing and relaxed state can determine the key nodes of abnormal causal connections between brain regions, and select the node with the largest difference in betweenness centrality as the priority stimulation site for intervention.

[0045] Step 2: Extract the instantaneous phases of individual Alpha and Theta oscillations through a phase estimation method.

[0046] As one embodiment of the present invention, the power spectral density of the Alpha and Theta frequency bands in the individual EEG signal is obtained using a spectral analysis method, and the phase change of the oscillating signal is estimated through an autoregressive model. Using these estimated phase information, the instantaneous phase of Alpha and Theta oscillations at each moment can be determined as the stimulation phase of otACS. By dynamically extracting the parameters of individual Alpha and Theta oscillations, the otACS stimulation is applied at the moment synchronized with the endogenous oscillations in the brain, ensuring the natural phase locking of external electrical stimulation and brain neural activities and optimizing the neuromodulation effect.

[0047] Step 3: Use spectral analysis to extract the resonance frequencies of individual Alpha oscillations and Theta oscillations as the stimulation frequencies of otACS.

[0048] It is understandable that the Alpha oscillation in the brain is usually associated with a relaxed state, with a frequency range of 8 - 12 Hz. The Theta oscillation is related to deeper mental states, such as meditation, dreams, and creative thinking processes, with a frequency range of 4 - 8 Hz. By extracting individualized brain wave characteristics (the resonance frequencies of Alpha oscillation and Theta oscillation), personalized and precise stimulation parameters can be formed according to the unique neural activity patterns of an individual, reflecting the natural frequencies of the brain in specific states, so that external electrical stimulation matches the endogenous oscillations of the brain. After extracting the individualized brain wave characteristics, these parameters (such as resonance frequency, instantaneous phase, etc.) are used to set the stimulation frequency and stimulation phase of otACS. Through precise frequency and phase adjustment, it is ensured that the external stimulation is synchronized with the natural oscillations of the brain, improving the intervention effect.

[0049] As one embodiment of the present invention, specifically, the resonance frequencies of Alpha oscillation and Theta oscillation are extracted through the following steps. Refer to Figure 3 , Figure 3 which is a flowchart for the extraction of the resonance frequencies of Alpha oscillation and Theta oscillation. First, the fast Fourier transform is used to calculate the power spectral curves of Alpha oscillation and Theta oscillation for each data segment. Due to differences between brain regions, the calculation is divided according to EEG leads; further, the spectral energy values of the individual Alpha frequency band and Theta frequency band are respectively superimposed at the data segment level; finally, the maximum value of the average power spectral curve at the corresponding data segment level is obtained as the resonance frequencies of Alpha oscillation and Theta oscillation.

[0050] As one embodiment of the present invention, the stimulation intensity of otACS is set between 0.1 and 10 mA. Specifically, the initial intensity is set in combination with the feedback of the wearer and dynamically adjusted according to the intervention effect during the subsequent intervention process.

[0051] After determining the above-mentioned personalized intervention parameters of otACS, the transcranial electrical stimulation execution module receives the personalized intervention parameters of otACS determined by the electrical stimulation control module, and performs personalized neural regulation on the subject through transcranial alternating current stimulation.

[0052] As one of the embodiments of the present invention, the transcranial alternating current stimulation execution module adopts a multi-channel transcranial alternating current stimulation device. The multi-channel transcranial alternating current stimulation device has channels covering the parietal lobe, temporal lobe, and prefrontal lobe, and regulates brain activities related to relaxation such as spatial perception, memory, learning, and emotion. The multi-channel transcranial alternating current stimulation device uses independent signal generators respectively. For the multi-channel transcranial alternating current stimulation device, the frequency range of the signal generator is 0.1 Hz - 1000 Hz, and the accuracy is 0.1 Hz. For the multi-channel transcranial alternating current stimulation device, the signal is output after passing through a power amplification circuit, and the stimulation intensity is 0.1 mA - 0.5 mA. For the multi-channel transcranial alternating current stimulation device, the electrode patches use circular carbon fiber electrode patches with a diameter of 10 mm, which are more easily attached to the scalp. The multi-channel transcranial alternating current stimulation device timely adjusts the output parameters of the signal generation circuit according to the preset current threshold and feedback signal to ensure that the stimulation current is within a safe and stable range. If an abnormal current (such as exceeding the set safety upper limit) is detected, the control unit will immediately cut off the output of the stimulation signal and issue an alarm prompt.

[0053] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Based on the disclosure and teachings of the above specification, those skilled in the art to which the present invention pertains can also make changes and modifications to the above embodiments. Therefore, the present invention is not limited to the specific embodiments disclosed and described above, and some modifications and changes to the present invention should also fall within the protection scope of the claims of the present invention. Based on the disclosure and teachings of the above specification, those skilled in the art to which the present invention pertains can also make changes and modifications to the above embodiments. Therefore, the present invention is not limited to the specific embodiments disclosed and described above, and some modifications and changes to the present invention should also fall within the protection scope of the claims of the present invention.

Claims

1. A transcranial electrical stimulation therapy system based on EEG-fNIRS feedback, characterized in that: The transcranial electrical stimulation therapy system comprises: An EEG-fNIRS signal acquisition and processing module is used to synchronously acquire and record EEG and fNIRS data of the subject in a soothing and relaxing state, and train a relaxation assessment model based on the EEG and fNIRS data; A relaxation assessment module, used for synchronously collecting and recording the real-time EEG and fNIRS data of the subject, performing relaxation assessment on the subject through the relaxation assessment model, and judging the relaxation state of the subject according to the result of the relaxation assessment; A transcranial electrical stimulation control module, which determines personalized intervention parameters of otACS based on the relaxation state of the subject; The transcranial electrical stimulation execution module receives the otACS personalized intervention parameters determined by the electrical stimulation control module, and performs personalized neural regulation on the subject through transcranial alternating current stimulation.

2. The transcranial electrical stimulation therapy system according to claim 1, characterized in that: The EEG-fNIRS signal acquisition and processing module includes an EEG-fNIRS signal acquisition module and an EEG-fNIRS signal processing module. The EEG signal acquisition module and the fNIRS signal acquisition module in the EEG-fNIRS signal acquisition module are spatially arranged at the same position.

3. The transcranial electrical stimulation therapy system according to claim 2, characterized in that: The EEG-fNIRS signal processing module is used to re-reference, filter, and remove artifacts from the collected EEG and fNIRS data, extract features related to relaxation, and then train a relaxation evaluation model based on a spatiotemporal feature convolutional network.

4. The transcranial electrical stimulation therapy system according to claim 3, characterized in that: The transcranial electrical stimulation control module determines the personalized intervention parameters of the otACS, including the stimulation frequency, stimulation phase, stimulation intensity and stimulation site of the otACS.

5. The transcranial electrical stimulation therapy system according to claim 3, characterized in that: The transcranial electrical stimulation control module determines the otACS stimulation site specifically including: constructing a brain function network model of the subject in a soothing and relaxing state based on causal analysis, extracting and recording the network betweenness centrality of the subject in the soothing and relaxing state; combining the real-time EEG and fNIRS data of the subject, constructing a brain function network model of the subject based on causal analysis, extracting the network betweenness centrality of the subject in real time, comparing the real-time determined network betweenness centrality with the network betweenness centrality of the subject in the soothing and relaxing state, and selecting the node with the largest difference in betweenness centrality as the priority stimulation site for intervention.

6. The transcranial electrical stimulation therapy system according to claim 3, characterized in that: The transcranial electrical stimulation control module determines the otACS stimulation frequency specifically by: extracting the resonance frequency of individual Alpha oscillations and Theta oscillations using spectrum analysis as the otACS stimulation frequency.

7. The transcranial electrical stimulation therapy system according to claim 3, characterized in that: The transcranial electrical stimulation control module extracts the resonant frequencies of the individual Alpha oscillation and Theta oscillation by the following steps: Fast Fourier transform was used to calculate the power spectrum curves of Alpha oscillation and Theta oscillation of each data segment. Due to the differences between brain regions, the calculation was divided according to EEG leads. The spectral energy values ​​of individual Alpha and Theta bands are superimposed at the data segment level; The maximum value of the average power spectrum curve at the corresponding data segment level is obtained as the resonant frequency of Alpha oscillation and Theta oscillation.

8. The transcranial electrical stimulation therapy system according to claim 3, characterized in that: The transcranial electrical stimulation control module determines the otACS stimulation phase specifically including: using phase estimation to extract the instantaneous phase of individual Alpha oscillation and Theta oscillation as the otACS stimulation phase.

9. The transcranial electrical stimulation therapy system according to claim 3, characterized in that: The transcranial electrical stimulation control module extracts the otACS stimulation phase through the following steps: using a spectrum analysis method to obtain the power spectral density of the Alpha and Theta frequency bands in the individual EEG signal, estimating the phase change of the oscillation signal through an autoregressive model; using the estimated phase information, determining the instantaneous phase of the Alpha and Theta oscillations at each moment.

10. The transcranial electrical stimulation therapy system according to claim 3, characterized in that: The spatiotemporal feature convolution network first uses a one-dimensional convolution kernel to adjust the EEG and fNIRS data to the same dimension, and then imports the spatiotemporal feature convolution network to extract EEG-fNIRS signal features; the spatiotemporal feature convolution network includes a time convolution kernel and a space convolution kernel, which respectively extract time dimension and space dimension features.

Images