Peripheral-central-peripheral evaluation system based on electroencephalogram-myoelectricity double-channel bidirectional mixed feedback

Abstract

The invention discloses a peripheral-central-peripheral evaluation system based on electroencephalogram-myoelectricity dual-channel bidirectional mixed feedback, and the system is characterized in that a muscle electrical stimulation module is used for providing a current with a specific frequency, and carrying out the electrical stimulation of muscles; the transcranial direct current stimulation module is used for providing current with specific frequency and performing electrical stimulation on the cerebral cortex; the myoelectricity acquisition module and the electroencephalogram acquisition module respectively acquire electric signal data of tested muscles and cerebral cortex and transmit the electric signal data to the upper computer feedback module, and the upper computer feedback module is used for controlling the muscle electrical stimulation module and the transcranial direct current stimulation module to generate electrical stimulation and acquiring data acquired by the myoelectricity acquisition module and the electroencephalogram acquisition module; and the data is evaluated after being processed, and an evaluation result is sent to the terminal display module for real-time display. According to the scheme, in the evaluation process, through a bidirectional mixed feedback mechanism, accurate positioning of cerebral cortex and muscle group stimulation targets is achieved, the electrical stimulation and evaluation effect can be optimized, quantitative evaluation is conducted on the result through a bidirectional evaluation model, scientific quantitative indexes are provided, the influence of individual differences and state changes on the evaluation result is reduced, and the evaluation accuracy is improved. And the reliability and stability of evaluation are improved.

Description

Technical Field

[0001] The present invention belongs to the field of neurorehabilitation assessment, and in particular relates to a peripheral-central-peripheral assessment system based on electroencephalogram-electromyography dual-channel bidirectional mixed feedback. Background Art

[0002] With the development of neuroscience and rehabilitation medicine, precision rehabilitation has become an important research direction. Traditional rehabilitation methods often rely on empirical judgment, which makes it difficult to accurately quantify the excitability of the cerebral cortex and the degree of muscle recovery, resulting in strong subjectivity in the evaluation of rehabilitation effects and large individual differences.

[0003] At present, some studies have used single EEG or sEMG signals for neurorehabilitation assessment, analyzing the brain activity state through EEG signals, but ignoring the direct feedback of muscle recovery; other studies have focused on the application of sEMG in muscle function assessment, and lacked consideration of the central regulatory mechanism of the brain.

[0004] Although the above methods each have their own advantages, none of them can truly achieve two-way closed-loop feedback between brain regions and muscle groups, cannot accurately locate specific key muscle groups and their recovery status, resulting in deviations in the evaluation of rehabilitation effects, and fail to fully utilize the complementarity of EEG and sEMG to improve the efficiency of neural stimulation. Summary of the invention

[0005] In response to the above problems, the purpose of the present invention is to provide a peripheral-central-peripheral assessment system based on EEG-EMG dual-channel bidirectional mixed feedback to improve the accuracy and reliability of neurorehabilitation assessment and achieve non-invasive, accurate and dynamic assessment of cerebral cortical excitability and specific muscle group recovery capacity.

[0006] The specific technical solution for achieving the purpose of the present invention is:

[0007] A peripheral-central-peripheral evaluation system based on EEG-EMG dual-channel bidirectional mixed feedback, comprising an EMG acquisition module, an EEG acquisition module, a muscle electrical stimulation module, a transcranial direct current stimulation module, a host computer feedback module, and a terminal display module;

[0008] The host computer feedback module is used to control the muscle electrical stimulation module and the transcranial direct current stimulation module to emit electrical stimulation, obtain the data collected by the electromyography acquisition module and the electroencephalography acquisition module, evaluate the data after processing, and send the evaluation results to the terminal display module for real-time display;

[0009] The muscle electrical stimulation module is used to provide a current of a specific frequency to electrically stimulate the muscles;

[0010] The transcranial direct current stimulation module is used to provide a current of a specific frequency to electrically stimulate the cerebral cortex;

[0011] The electromyography acquisition module and the electroencephalography acquisition module respectively acquire the electrical signal data of the subject's muscles and cerebral cortex, and transmit the data to the host computer feedback module.

[0012] The terminal display module is used to visually display the evaluation process and results.

[0013] Furthermore, the electromyography acquisition module acquires grip strength data and muscle electrical signal data;

[0014] The EEG acquisition module acquires EEG signal data.

[0015] Furthermore, the process in which the host computer feedback module controls the muscle electrical stimulation module and the transcranial direct current stimulation module to emit electrical stimulation and obtains the data collected by the electromyography acquisition module and the electroencephalography acquisition module is as follows:

[0016] Obtain grip strength data of the subject's muscles, obtain myoelectric data of a wide range of muscle groups during the gripping process, and obtain EEG data of the cerebral cortex when the muscle groups are in the best state;

[0017] Determine the active brain area according to the EEG data of the cerebral cortex obtained above as the stimulation target area of ​​transcranial direct current stimulation, and control the transcranial direct current stimulation module to electrically stimulate the stimulation target area;

[0018] Acquire data from the electromyography acquisition module and the electroencephalography acquisition module, and determine the muscle group with the greatest electrical signal feedback intensity and the fastest reaction speed according to the collected electromyography data as the target muscle group;

[0019] The transcranial direct current stimulation module is controlled to electrically stimulate the target stimulation area, and the muscle electrical signals fed back by the target muscle group and the EEG signals of the target area are synchronously obtained during the process. The optimal intensity range of transcranial direct current stimulation is selected according to the optimal state of the mixed feedback of the muscle electrical signals.

[0020] Furthermore, after determining the target area, the subject's muscles are again required to perform a grip operation, and the optimal grip value and the average state value of the target muscle group are determined based on the real-time curve feedback of the muscle electrical signal of the target muscle group to determine the accuracy of the target area selection.

[0021] Furthermore, the transcranial direct current stimulation module electrically stimulates the target area for a certain period of time according to the determined optimal intensity interval, the muscle electrical stimulation module electrically stimulates the target muscle group, and the host computer feedback module processes the muscle electrical signal data and EEG signal data of the target muscle group using a two-way evaluation model, and fuses and compares the data with the data obtained before electrical stimulation to obtain an evaluation result.

[0022] Furthermore, the bidirectional evaluation model is constructed based on the Transformer model and a deep convolutional neural network, wherein the deep convolutional neural network is used to extract the spatial and temporal features of the muscle electrical signal data and the electroencephalogram signal data, and the Transformer model is used to capture the long-term dependencies in the time series data.

[0023] Furthermore, the bidirectional evaluation model includes a data preprocessing layer, a feature extraction layer, an attention mechanism layer, a bidirectional LSTM layer, and an evaluation output layer;

[0024] The data preprocessing layer is used to perform standardization processing and sliding window segmentation processing on the muscle electrical signal data and the brain electrical signal data;

[0025] The feature extraction layer extracts deep features of muscle electrical signal data and EEG signal data through multi-layer convolution and pooling;

[0026] The attention mechanism layer uses a self-attention mechanism to capture the long-term dependencies and interrelationships of extracted features in time series;

[0027] The bidirectional LSTM layer is used to capture the bidirectional information of the time series;

[0028] The evaluation output layer is used to flatten the output of the bidirectional LSTM and map it to the final feedback sensitivity value through one or two fully connected layers to obtain the evaluation value.

[0029] Furthermore, the standardization processing of the data preprocessing layer is to standardize the muscle electrical signal data and the brain electrical signal data signals after the optimization and preprocessing so that they have the same mean and variance;

[0030] The sliding window segmentation process is to segment the standardized muscle electrical signal data and brain electrical signal data into multiple shorter segments, and process each segment as a subsequent sample.

[0031] Furthermore, the tested muscle group includes the biceps brachii, brachialis, and triceps brachii of the upper limb and the pronator teres, radial flexor carpi, palmaris longus, ulnar flexor carpi, pronator quadratus, radial extensor carpi longus, radial extensor carpi brevis, ulnar extensor carpi, and supinator of the forearm.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] (1) The solution of the present invention uses a two-way mixed feedback mechanism to achieve precise positioning of the stimulation targets of the cerebral cortex and muscle groups during the evaluation process, thereby improving the accuracy and reliability of the evaluation; allowing the user to dynamically adjust the stimulation intensity according to the feedback results to optimize the electrical stimulation and evaluation effects;

[0034] (2) The solution of the present invention uses a two-way evaluation model to quantify the results, provide scientific quantitative indicators, reduce the impact of individual differences and state changes on the evaluation results, and improve the reliability and stability of the evaluation.

[0035] (3) The evaluation of the present invention is non-invasive and safe. Non-invasive means are used throughout the evaluation process, which reduces user pain, improves user experience, and ensures the safety of the use process.

[0036] The present invention is further described below in conjunction with specific implementation modes. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of the architecture of the peripheral-central-peripheral assessment system based on EEG-EMG dual-channel bidirectional mixed feedback of the present invention.

[0038] Figure 2 It is a schematic diagram of bidirectional mixed feedback of the peripheral-central-peripheral evaluation system based on EEG-EMG dual-channel bidirectional mixed feedback of the present invention.

[0039] Figure 3 It is an implementation flow chart of the peripheral-central-peripheral assessment based on EEG-EMG dual-channel bidirectional mixed feedback of the present invention.

[0040] Figure 4 It is a schematic diagram of the bidirectional evaluation model architecture of the present invention.

[0041] Figure 5 Schematic diagram of a transcranial direct current stimulation module in an embodiment of the present invention.

[0042] Figure 6 Schematic diagram of a muscle electrical stimulation module in an embodiment of the present invention.

[0043] Figure 7 Schematic diagram of a muscle group electrode patch in an embodiment of the present invention. DETAILED DESCRIPTION

[0044] Combination Figure 1 , a peripheral-central-peripheral evaluation system based on EEG-EMG dual-channel bidirectional mixed feedback, including an EMG acquisition module, an EEG acquisition module, a muscle electrical stimulation module, a transcranial direct current stimulation module, a host computer feedback module and a terminal display module;

[0045] The host computer feedback module is used to control the muscle electrical stimulation module and the transcranial direct current stimulation module to emit electrical stimulation, obtain the data collected by the electromyography acquisition module and the electroencephalography acquisition module, evaluate the data after processing, and send the evaluation results to the terminal display module for real-time display;

[0046] The muscle electrical stimulation module is used to provide a specific low-frequency current to electrically stimulate the muscles to cause them to twitch or contract. The muscle electrical stimulation module of this solution is a multi-channel neuromuscular electrical stimulator, which generally provides a 20-50Hz low-frequency current;

[0047] The transcranial direct current stimulation module is used to provide a current of a specific frequency, generally a low-intensity direct current (1-2 mA), to electrically stimulate the cerebral cortex to regulate the activity of neurons in the cerebral cortex;

[0048] The electromyography acquisition module and the electroencephalography acquisition module respectively acquire the electrical signal data of the subject's muscles and cerebral cortex, and transmit the data to the host computer feedback module.

[0049] The electromyography acquisition module of the present solution includes a grip force sensor and a multi-channel surface muscle electrical signal acquisition device, wherein the multi-channel surface muscle electrical signal acquisition device acquires electrical signal data of the subject's muscles during the entire evaluation process; the electroencephalogram acquisition module acquires electroencephalogram signal data.

[0050] The terminal display module is used to visually display the evaluation process and results.

[0051] Combination Figure 2 and Figure 3 The process in which the host computer feedback module controls the muscle electrical stimulation module and the transcranial direct current stimulation module to emit electrical stimulation and obtains the data collected by the electromyography acquisition module and the electroencephalography acquisition module is as follows:

[0052] First, the grip strength data of the tested muscles were obtained;

[0053] The multi-channel neuromuscular electrical stimulator is widely attached to the biceps, brachialis, and triceps of the upper arm of the subject's upper limb, as well as the pronator teres, radial flexor carpi, palmaris longus, ulnar flexor carpi, pronator quadratus, radial extensor carpi longus, radial extensor carpi brevis, ulnar extensor carpi, supinator and other muscle groups of the forearm, and the hand is clenched to the maximum value of the grip sensor to obtain the maximum grip value of the subject's muscle. At this time, the electromyographic acquisition module is controlled to obtain the electromyographic data of a wide range of muscle groups of the subject's muscles. Here, the average value of the patch muscle group is taken, and the optimal state value of the initial grip force and the average state value of the patch muscle group can be read through the real-time curve feedback of the muscle electrical signal displayed on the terminal, and the electroencephalographic data of the cerebral cortex at this time are obtained at the same time;

[0054] When the upper limb muscle energy of the subject is in the best state, the EEG images of all brain regions of the subject are read in real time. According to the EEG data of the cerebral cortex, the image segmentation algorithm model is used to determine the active brain area as the stimulation target area of ​​the transcranial direct current stimulation, and the transcranial direct current stimulation module is controlled to electrically stimulate the stimulation target area. Generally, when the arms of the subject are completely relaxed, the strongly associated hot spot area is given an electrical stimulation of 1-2 mA with a gradient difference of 0.2 mA for a duration of 10 minutes;

[0055] While conducting electrical stimulation, data from the electromyography acquisition module and the electroencephalography acquisition module are acquired, and the muscle group with the largest electrical signal feedback intensity and the fastest reaction speed is determined based on the acquired electromyography data, and the target muscle group is determined. Generally, the top six muscle groups with the largest average electrical signal feedback intensity and the fastest reaction speed per unit time are selected to achieve accurate positioning of the most effective target muscle group for muscle electrical stimulation;

[0056] Afterwards, the subjects' muscles were again required to perform the same grip operation as in the above process, and the multi-channel electromyographic electrodes were accurately attached to the target muscle group selected by the subject. The terminal displayed the real-time curve feedback of the sEMG muscle electrical signal to read the best grip strength value and the average state value of the target muscle group, so as to judge the accuracy and effectiveness of the target area selection.

[0057] Then, the transcranial direct current stimulation module is controlled to electrically stimulate the stimulation target area. For example, when the subjects' arms are completely relaxed, 1-2 mA of electrical stimulation is continuously increased for 10 minutes to the brain stimulation target. The muscle electrical signals and brain electrical signals fed back in the process are synchronously obtained, and the optimal intensity range of the transcranial direct current stimulation is selected according to the optimal state of the mixed feedback of the muscle electrical signals.

[0058] Finally, after the subject performed 20 minutes of low-frequency electrical stimulation of the target muscle group and transcranial direct current stimulation of the optimal intensity of the stimulation target, the host computer feedback module was used to perform a two-way evaluation algorithm model on the subject's EEG and sEMG dual-channel signals. After fusion and comparison with the subject's muscle electrical stimulation and transcranial direct current data, the effect of this electrical stimulation was evaluated, and relevant comparative visualization charts were given through the terminal display.

[0059] In addition, combined Figure 4 The bidirectional evaluation model described in this scheme is constructed based on the Transformer model and the deep convolutional neural network, in which the deep convolutional neural network is used to extract the spatial and temporal features of the muscle electrical signal data and the EEG signal data, and the Transformer model is used to capture the long-term dependencies in the time series data.

[0060] The bidirectional evaluation model includes a data preprocessing layer, a feature extraction layer, an attention mechanism layer, a bidirectional LSTM layer and an evaluation output layer;

[0061] The data preprocessing layer is used to perform standardization processing and sliding window segmentation processing on the muscle electrical signal data and the brain electrical signal data;

[0062] The standardization processing of the data preprocessing layer is to standardize the muscle electrical signal data and the brain electrical signal data signals after optimization and preprocessing so that they have the same mean and variance;

[0063] The sliding window segmentation process is to segment the standardized muscle electrical signal data and the brain electrical signal data into multiple shorter segments, and each segment is used as a subsequent sample for processing. Taking the 10-minute evaluation process as an example, the continuous signal can be segmented into multiple shorter segments (such as 1 second or shorter), and each segment is processed as a sample;

[0064] The feature extraction layer extracts deep features of muscle electrical signal data and EEG signal data through multi-layer convolution and pooling;

[0065] Generally speaking, including convolutional layers: Use multiple convolutional layers (e.g., 3-5 layers) to extract the spatial and temporal features of the signal. Each layer contains multiple convolution kernels (such as 32, 64, 128, etc.) to capture features of different scales.

[0066] Activation function: ReLU activation function is used after each convolution layer to increase nonlinearity.

[0067] Pooling layer: A maximum pooling layer is added between every two convolutional layers to reduce the feature dimension and introduce translation invariance.

[0068] The attention mechanism layer uses the self-attention mechanism to capture the long-term dependencies and interrelationships of extracted features in time series, which generally includes:

[0069] Position encoding: Since the Transformer itself does not have position information, it is necessary to add position encoding to the segmented fragments.

[0070]

[0071] Among them, pos is the position index, i is the dimension index, and dmodel is the embedding dimension of the model.

[0072] Self-attention mechanism: The feature sequence extracted by DCNN is input into the encoder part of Transformer, and the self-attention mechanism is used to capture the dependencies between elements in the sequence.

[0073]

[0074] where Q, K, and V are matrices of query, key, and value, respectively. k is the dimension of the key vector.

[0075] Multi-head attention: Using a multi-head attention mechanism allows the model to learn interactions of information from different subspaces simultaneously.

[0076] textMultiHead(Q,K,V)=Concat(head1,head2,...,headh )W O

[0077] texthead i =Attention(QW i Q ,KW i K ,VW i V )

[0078] W i Q , W i K , W i V is the projection matrix of the i-th head, i is the output projection matrix, W O is the output projection matrix.

[0079] The bidirectional LSTM layer is used to capture the bidirectional information of the time series;

[0080] The evaluation output layer is used to flatten the output of the bidirectional LSTM and map it to the final feedback sensitivity value through one or two fully connected layers to obtain the evaluation value.

[0081] Example

[0082] In order to clearly describe the technical solution and the effects achieved by the present invention, the technical solution of the present invention is clearly explained below in combination with the accompanying drawings and specific embodiments of the present invention, so that technical personnel in the field can implement the invention without performing creative work. The structure shown in the accompanying drawings is not the entire actual structure but only a part of the actual structure. It should be noted that all other embodiments made by ordinary technical personnel in the field on the basis of the embodiments of the present invention without performing creative work should fall within the protection scope of the present invention.

[0083] The following examples are only for better explanation of the present invention, but the content of the present invention is not limited to the application in the examples. Therefore, those skilled in the art can make non-essential modifications to the implementation scheme based on the above invention content and apply it to other embodiments, which is still within the protection scope of the present invention. In addition, the experimental methods without specific conditions in the following examples should be carried out according to the conventional conditions or the conditions recommended by the manufacturer. If the expressions in the text are not specially explained, they are only used for distinction and have no other meanings.

[0084] Combination Figures 5 to 7In this embodiment, the subject lets his arms hang naturally on both sides of the body, and the multi-conductor electromyographic electrodes are widely attached to the biceps, brachialis, and triceps of the subject's upper limbs, as well as the pronator teres, radial flexor carpi, palmaris longus, ulnar flexor carpi, pronator quadratus, radial extensor carpi longus, radial extensor carpi brevis, ulnar extensor carpi, supinator and other muscle groups of the forearm. The affected hand holds the grip force sensor to the maximum value, and the subject's maximum grip force value is tested. The optimal initial grip force state value and the average state value of the above-mentioned patch muscle group are read through the screen sEMG muscle electrical signal real-time curve feedback.

[0085] When the upper limb muscle energy of the subject is in the optimal state, the EEG images of all brain regions of the subject are read in real time. The host computer uses the image segmentation algorithm model to find the most active brain region as the stimulation target area of ​​​​transcranial direct current stimulation.

[0086] When the subjects' arms were completely relaxed, the strongly associated hot spot area was given an electrical stimulation of 1-2 mA with a gradient difference of 0.2 mA for 10 minutes.

[0087] While electrical stimulation is being performed, the host computer uses a two-way hybrid algorithm model to read and analyze the electrical signal feedback intensity of the above-mentioned muscle groups in real time, and selects the top six muscle groups with the largest average electrical signal feedback intensity and the fastest reaction speed per unit time, thereby achieving accurate positioning of the most effective target muscle groups for muscle electrical stimulation.

[0088] The subjects were again asked to perform the same gripping operation as in the above process, and to accurately attach the multi-conductor electromyographic electrodes to the target muscle groups selected by the subjects. The best gripping strength value and the average state value of the target muscle group were read through the real-time curve feedback of the sEMG muscle electrical signal on the screen to determine the accuracy and effectiveness of the target area selection.

[0089] Afterwards, with the subjects' arms completely relaxed, the brain stimulation target was given an electrical stimulation of 1 to 2 mA continuously for 10 minutes, and the muscle electrical signals and EEG signals of the target area were synchronously transmitted to the host computer. During each transcranial direct current stimulation, the host computer selected the optimal intensity range of transcranial direct current stimulation based on the optimal state of mixed feedback of muscle electrical signals through a two-way hybrid algorithm model.

[0090] After the subjects performed 20 minutes of low-frequency electrical stimulation of the target muscle groups and transcranial direct current stimulation of the optimal intensity of the stimulation target, the host computer used an algorithm model to mix and process the subject's EEG and sEMG dual-channel signals. After fusion and comparison with the subject's data before electrical stimulation, the effect of this electrical stimulation was evaluated, and relevant comparative visualization line statistical charts and further processing suggestions were given through the terminal APP.

[0091] The solution of the present invention uses a two-way mixed feedback mechanism during the evaluation process to achieve precise positioning of the stimulation targets of the cerebral cortex and muscle groups, so as to improve the accuracy and reliability of electrical stimulation and evaluation; it enables the user to dynamically adjust the stimulation intensity according to the feedback results to optimize the electrical stimulation and evaluation effects; at the same time, the results are quantitatively evaluated through a two-way evaluation model to provide scientific quantitative indicators, reduce the impact of individual differences and state changes on the evaluation results, and improve the reliability and stability of the evaluation. In addition, the evaluation of the present invention is non-invasive and safe, and non-invasive means are used throughout the evaluation process, which reduces user pain, improves user experience, and ensures the safety of the use process.

[0092] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A peripheral-central-peripheral assessment system based on EEG-EMG dual-channel bidirectional mixed feedback, characterized in that: It includes electromyography acquisition module, electroencephalography acquisition module, muscle electrical stimulation module, transcranial direct current stimulation module, host computer feedback module and terminal display module; The host computer feedback module is used to control the muscle electrical stimulation module and the transcranial direct current stimulation module to emit electrical stimulation, obtain the data collected by the electromyography acquisition module and the electroencephalography acquisition module, evaluate the data after processing, and send the evaluation results to the terminal display module for real-time display; The muscle electrical stimulation module is used to provide a current of a specific frequency to electrically stimulate the muscles; The transcranial direct current stimulation module is used to provide a current of a specific frequency to electrically stimulate the cerebral cortex; The electromyography acquisition module and the electroencephalography acquisition module respectively collect electrical signal data of the subject's muscles and cerebral cortex, and transmit them to the host computer feedback module; The terminal display module is used to visually display the evaluation process and results.

2. The peripheral-central-peripheral assessment system based on EEG-EMG dual-channel bidirectional mixed feedback according to claim 1 is characterized in that: The electromyography acquisition module acquires grip strength data and muscle electrical signal data; The EEG acquisition module acquires EEG signal data.

3. The peripheral-central-peripheral assessment system based on EEG-EMG dual-channel bidirectional mixed feedback according to claim 1, characterized in that: The process in which the host computer feedback module controls the muscle electrical stimulation module and the transcranial direct current stimulation module to emit electrical stimulation and obtains the data collected by the myoelectric acquisition module and the electroencephalographic acquisition module is as follows: Obtain grip strength data of the subject's muscles, obtain myoelectric data of a wide range of muscle groups during the gripping process, and obtain EEG data of the cerebral cortex when the muscle groups are in the best state; Determine the active brain area according to the EEG data of the cerebral cortex obtained above as the stimulation target area of ​​transcranial direct current stimulation, and control the transcranial direct current stimulation module to electrically stimulate the stimulation target area; Acquire data from the electromyography acquisition module and the electroencephalography acquisition module, and determine the muscle group with the greatest electrical signal feedback intensity and the fastest reaction speed according to the collected electromyography data as the target muscle group; The transcranial direct current stimulation module is controlled to electrically stimulate the target stimulation area, and the muscle electrical signals fed back by the target muscle group and the EEG signals of the target area are synchronously obtained during the process. The optimal intensity range of transcranial direct current stimulation is selected according to the optimal state of the mixed feedback of the muscle electrical signals.

4. The peripheral-central-peripheral assessment system based on EEG-EMG dual-channel bidirectional mixed feedback according to claim 3 is characterized in that: After determining the target area, the subjects were asked to perform the grip operation again, and the optimal grip value and the average state value of the target muscle group were determined based on the real-time curve feedback of the muscle electrical signals of the target muscle group to judge the accuracy and effectiveness of the target area selection.

5. The peripheral-central-peripheral assessment system based on EEG-EMG dual-channel bidirectional mixed feedback according to claim 3 is characterized in that: The transcranial direct current stimulation module electrically stimulates the target area for a certain period of time according to the determined optimal intensity interval, the muscle electrical stimulation module electrically stimulates the target muscle group, and the host computer feedback module processes the muscle electrical signal data and EEG signal data of the target muscle group using a two-way evaluation model, and fuses and compares the data obtained before electrical stimulation to obtain an evaluation result.

6. The peripheral-central-peripheral assessment system based on EEG-EMG dual-channel bidirectional mixed feedback according to claim 5, characterized in that: The bidirectional evaluation model is constructed based on the Transformer model and a deep convolutional neural network, wherein the deep convolutional neural network is used to extract the spatial and temporal features of muscle electrical signal data and electroencephalogram signal data, and the Transformer model is used to capture the long-term dependencies in time series data.

7. The peripheral-central-peripheral assessment system based on EEG-EMG dual-channel bidirectional mixed feedback according to claim 5, characterized in that: The bidirectional evaluation model includes a data preprocessing layer, a feature extraction layer, an attention mechanism layer, a bidirectional LSTM layer and an evaluation output layer; The data preprocessing layer is used to perform standardization processing and sliding window segmentation processing on the muscle electrical signal data and the brain electrical signal data; The feature extraction layer extracts deep features of muscle electrical signal data and EEG signal data through multi-layer convolution and pooling; The attention mechanism layer uses a self-attention mechanism to capture the long-term dependencies and interrelationships of extracted features in time series; The bidirectional LSTM layer is used to capture the bidirectional information of the time series; The evaluation output layer is used to flatten the output of the bidirectional LSTM and map it to the final feedback sensitivity value through one or two fully connected layers to obtain the evaluation value.

8. The peripheral-central-peripheral assessment system based on EEG-EMG dual-channel bidirectional mixed feedback according to claim 7, characterized in that: The standardization processing of the data preprocessing layer is to standardize the muscle electrical signal data and the brain electrical signal data signals after optimization and preprocessing so that they have the same mean and variance; The sliding window segmentation process is to segment the standardized muscle electrical signal data and brain electrical signal data into multiple shorter segments, and process each segment as a subsequent sample.

9. The peripheral-central-peripheral assessment system based on EEG-EMG dual-channel bidirectional mixed feedback according to claim 2, characterized in that: The tested muscle groups include the biceps brachii, brachialis, and triceps brachii of the upper limb and the pronator teres, flexor carpi radialis, palmaris longus, flexor carpi ulnaris, pronator quadratus, extensor carpi radialis longus, extensor carpi radialis brevis, extensor carpi ulnaris, and supinator of the forearm.

Images