CN110495880B

CN110495880B - Movement disorder cortical plasticity management method based on transcranial electrical stimulation brain muscle coupling

Info

Publication number: CN110495880B
Application number: CN201910758648.8A
Authority: CN
Inventors: 陆晟; 罗志增; 席旭刚; 马存斌
Original assignee: Hangzhou Dianzi University
Current assignee: Hangzhou Dianzi University
Priority date: 2019-08-16
Filing date: 2019-08-16
Publication date: 2022-04-08
Anticipated expiration: 2039-08-16
Also published as: CN110495880A

Abstract

The invention provides a movement disorder cortical plasticity management method based on transcranial electrical stimulation brain muscle coupling. The invention firstly analyzes the multi-lead EEG signal of the functional brain area, calculates the brain function network index of the communication rate and the small world characteristic by adopting a brain function network modeling method, compares the brain function network characteristics before and after tDCS stimulation and reveals the function connection and nerve remodeling rule of the functional forezone, the functional area and the function perception area of the stimulated hemisphere during the movement activity. Then, a multi-level neuromuscular coupling analysis method based on the causal measurement of the dynamic regression model is adopted to extract brain-brain, brain-muscle and muscle-muscle coupling characteristic indexes, and the rehabilitation effect of the upper limb motor function is described from different sides. And finally, researching the correlation between the brain-muscle electrical coupling characteristics and the influence rule of tDCS on the neural plasticity, providing a basis for further improving the stimulation mode and parameters of the tDCS, and realizing effective guidance and management of motor function cortical nerve remodeling.

Description

Movement disorder cortical plasticity management method based on transcranial electrical stimulation brain muscle coupling

Technical Field

The invention belongs to the field of pattern recognition, and relates to a motor rehabilitation training method by utilizing motor imagery in an active rehabilitation robot, a motor dysfunction cortex plasticity management method based on external stimulation intervention and motor rehabilitation evaluation, in particular to a motor dysfunction cortex plasticity management method based on transcranial direct current stimulation and brain and muscle electrical coupling analysis.

Background

Cerebral apoplexy is one of common cerebrovascular diseases in the world, and patients often leave motor dysfunction after the disease, and the motor function is slow to recover, difficult and poor in prognosis, so that the cerebral apoplexy becomes a key point and a difficult point in rehabilitation treatment. In recent years, the rapid development of brain science and the interdisciplinary research results of multiple disciplines such as neurophysiology, brain function network theory, brain-computer interface and the like bring a new means for the motor rehabilitation of stroke patients, and become an important way for improving the rehabilitation effect and relieving the shortage of rehabilitation resources.

The neural plasticity theory indicates that the patients with damaged central nervous system adopt timely and reasonable autonomic rehabilitation therapy, and can rebuild motor nerve pathways through the change of the nerve tissue form or compensation, so that the motor function is recovered to a certain extent. In recent years, transcranial electrical stimulation (tDCS) has attracted wide attention as a central intervention rehabilitation therapy method, mainly by stimulating related functional brain regions, regulating cortical activity, improving synaptic plasticity, and causing irreversible changes in plasticity caused by long-term effects. While tDCS has clear advantages in directly regulating cortical activity and directing motor impulses to the spinal cord tract, there is a lack of understanding of the laws of their effects. The brain function network analysis method can reflect the information transmission and division work cooperation mechanism of each brain area when limbs move from the global perspective, and can effectively express the movement law of the kinesthetic cortex. Therefore, the invention adopts a complex brain function network modeling method, expresses the topological structure of the brain function network by brain electrical signal incidence matrixes among different brain areas, extracts topological characteristics describing the relation among network nodes by methods such as graph theory, spectrum analysis and matrix theory and the like, and expresses the influence rule of tDCS on cortical activity and brain function network characteristics.

At present, most rehabilitation treatment methods based on tDCS adopt fixed stimulation modes and parameters, and a flexible change mechanism is not made according to rehabilitation effect evaluation, so that the remodeling effect of motor function central nerves is restricted. According to the theory of human motor neurophysiology, normal motor control means that the central nervous system uses the existing and past information to convert the nerve energy into kinetic energy through the participation of skeletal muscles and complete effective functional activities. The electroencephalogram (EEG) and the Electromyogram (EMG) signals respectively represent the motion control information of the brain to the muscle and the sensory feedback information activity state of the muscle function response, and the functional relation between the nerve and the muscle in the motion control process can be reflected through brain and electromyogram coupling analysis, so that a theoretical basis is provided for understanding the motion control process and the pathological mechanism of dyskinesia, and a reliable biomarker can be provided for evaluating the rehabilitation effect of the motor function. The invention researches the influence rule of tDCS on neural plasticity by utilizing a multilevel brain function network and cortical muscle coupling analysis technology based on EEG and EMG, carries out timely and comprehensive motor function evaluation on the aspect of rehabilitation effect, and adjusts the tDCS intervention mode according to the evaluation result to realize effective management of neural remodeling.

Disclosure of Invention

In order to overcome the defects of the existing rehabilitation treatment method and improve the remodeling effect of the motor function central nerve, the invention provides a motor function disorder cortical plasticity management method based on tDCS and brain-muscle electrical coupling analysis. Firstly, analyzing functional brain area multi-lead EEG signals, calculating brain function network indexes of a communication rate and characteristics of a small world by adopting a brain function network modeling method, and comparing the characteristics of the brain function network before and after tDCS stimulation so as to reveal the characteristic change of the brain function network connection relation of a functional anterior area, a functional area and a functional sensing area of a stimulated hemisphere during the movement activity. And then extracting the characteristic indexes of brain-brain and brain-muscle coupling by adopting a multi-level neuromuscular coupling analysis method based on the causal measurement of the dynamic regression model to obtain the rehabilitation effect. And finally, researching the correlation between the brain-muscle electrical coupling characteristics and the influence rule of tDCS electrical stimulation on the neural plasticity, providing a basis for further improvement of the stimulation mode and parameters of the tDCS, and realizing effective guidance and management of motor function cortical nerve remodeling.

In order to achieve the above object, the method of the present invention mainly comprises the following steps:

step 1, collecting EEG data and corresponding part electromyogram EMG data of a patient in different rehabilitation training stages.

Step 2, extracting brain function network characteristics reflecting nerve activity characteristics by adopting a brain function network modeling method, revealing the influence rule of transcranial direct current stimulation tDCS on the sensory-motor cortex activeness and the central nerve plasticity change, and specifically comprising the following steps:

2-1, establishing a brain function network correlation matrix. The measured area of each lead defining the EEG is a node of the network whose electrical activity is in time series, and the correlation matrix is established by establishing the connection between each pair of EEG signals of each channel and calculating their strengths.

2-2, establishing a adjacency matrix and a brain function network topology. And on the basis of the correlation matrix, converting the correlation matrix into a sparse adjacent matrix through threshold processing, and judging whether a connecting edge exists between an element value of the adjacent matrix and a brain region node or not, thereby constructing a brain function network topology.

And 2-3, extracting the characteristics of the brain function network. The feature extraction method is to research the relationship between network features and central nervous plasticity indexes by using graph theory. The neural plasticity index refers to the change of the brain function network, is also measured by the characteristics of the connection relation of the brain function network, and is based on the scale judgment result of the motor function rehabilitation. And (4) selecting the characteristic path length, the clustering coefficient and the betweenness of the directed network to describe the brain function network characteristics of the central nervous plasticity change under the tDCS.

And 2-4. carrying out correlation analysis on the motor cortex brain function network characteristics under tDCS. Analyzing the multi-lead EEG signals by adopting a brain function network, drawing undirected graphs of all frequency bands of the EEG according to the correlation matrix obtained by 2-1, and calculating the index of the brain function network: and comparing the brain function network indexes before and after the tDCS stimulation so as to reveal the characteristic change of the brain function network connection relation of the premotor, the motor area and the motion perception area of the stimulated hemisphere during the motion activity.

Step 3, extracting brain-brain, brain-muscle and muscle-muscle coupling characteristic indexes from the data collected in the step 1 by adopting a multi-level neuromuscular coupling analysis method based on the causal measurement of a dynamic regression model, and evaluating the rehabilitation effect of the exercise function, wherein the method comprises the following specific steps:

and 3-1, respectively preprocessing the EEG and EMG signals. Aiming at the EEG signal, firstly, the EEG signal is subjected to band-pass filtering of 0.1-100Hz, meanwhile, noise interference of baseline drift and power frequency is filtered, and then, the artifact in the EEG signal is eliminated by adopting a reference independent variable analysis method in combination with the prior knowledge of the ocular artifact. Then, according to the state of the study object and the subject in the experiment, corresponding frequency band signals in each channel of the EEG, such as 1-4 Hz delta wave, 4-8 Hz theta wave and 8-12 Hz alpha wave, are extracted. Aiming at the EMG signals, a 50Hz wave trap is adopted to filter power frequency interference in the EMG signals, and then 0.1-100Hz band-pass filtering is carried out.

And 3-2, optimizing and selecting the electroencephalogram electrodes related to the upper limb movement by adopting a typical correlation analysis CCA theory to obtain an optimal electroencephalogram lead set. Selecting the correlation degree of the EEG signal and the EMG signal during the upper limb movement by adopting a CCA method, and selecting an electroencephalogram electrode with higher correlation degree with the specific upper limb movement as an optimal electroencephalogram lead set for subsequent electroencephalogram and electromyogram coupling analysis by taking a CCA coefficient as a reference basis.

And 3-3, calculating causal values among the multi-lead EEG-EEG, EEG-EMG and EMG-EMG signals by using causal measures based on the dynamic regression model. The method comprises the steps of firstly extracting different frequency bands from a time sequence by adopting a wavelet decomposition method and a multivariate empirical mode decomposition method, and then calculating causal values among multi-lead signals under each frequency band.

And 3-4, describing the progress change of the recovery of the upper limb motor function by the brain-brain, brain-muscle and muscle-muscle coupling characteristic indexes, and performing rehabilitation evaluation by a clinician by using a Fugl-Meyer upper limb motor function evaluation scale.

And 4, researching the correlation between indexes such as brain-brain, brain-muscle and muscle-muscle coupling characteristics and the like in the step 3 and the influence rule of electrical stimulation on the neural plasticity, and guiding the stimulation mode and parameters of the tDCS in the next step to achieve plasticity management of motor cortex rehabilitation.

The invention describes the cortex activity and plasticity change by brain function network characteristics, describes the exercise rehabilitation effect by brain-muscle electrical coupling characteristics, establishes the corresponding relation between tDCS stimulation and the cortex activity and plasticity change, and adjusts the tDCS stimulation to achieve the purpose of managing the cortex plasticity through exercise rehabilitation evaluation.

Drawings

FIG. 1 is a functional block diagram of an implementation of the present invention;

FIG. 2 is a flow chart of a study of the effects of tDCS on kinesthetic cortical activity in an embodiment of the present invention;

Detailed Description

The embodiments of the present invention will be described in detail below with reference to the accompanying drawings: the embodiment is implemented on the premise of the technical scheme of the invention, and a detailed implementation mode and a specific operation process are given.

As shown in fig. 1, the present embodiment includes the following steps:

step one, building a rehabilitation experiment platform, determining an experiment paradigm and formulating a rehabilitation training scheme, wherein the specific process is as follows:

the method comprises the steps of adopting a tDCS device to form an electrical stimulation intervention channel of cerebral cortex, and utilizing a multi-channel EEG signal acquisition function of the tDCS device and a wireless EMG signal acquisition device to respectively form EEG and EMG signal acquisition channels.

The experimental subjects were 40 post-stroke upper limb dysfunction patients who met the inclusion criteria. After the severity of upper limb dysfunction of a patient is firstly evaluated, an electrical stimulation intervention scheme is formulated according to the specific condition of the patient. The protocol defines a short-term test cycle for one week and a long-term test cycle for one month, 5 days per week, once a day. The experimental protocol was as follows: selecting specific upper limb movements which are relatively complete in medical examination and evaluation method as target movements of electrical stimulation intervention, wherein the specific upper limb movements comprise basic movements of elbow flexion, elbow extension, wrist flexion, wrist extension, fist making and fist unfolding. The subject is first given a 20 minute tDCS and then stationary EEG signals and active EMG signals are acquired.

Step two, extracting brain function network characteristics reflecting the characteristics of neural activity by adopting a brain function network modeling method, and revealing the influence rule of tDCS on the sensory-motor cortex activity and the central nervous plasticity change, wherein the specific process is as follows:

the law of the tDCS on the motor sensory cortex activity is researched and analyzed by adopting a brain function network modeling method, and the research steps are shown in figure 2.

Step 1: and establishing a brain function network correlation matrix. Establishing a connection relation between every two EEG signals of each channel by adopting quantization methods of cross correlation, mutual information quantity, phase synchronization, partial directional coherent analysis and a synchronous likelihood method, and calculating intensity values of the EEG signals, thereby establishing a correlation matrix;

step 2: and establishing an adjacency matrix and brain function network topology. On the basis of the correlation matrix, the correlation matrix is converted into a sparse adjacent matrix by taking weak connecting edges removed, network connectivity guaranteed and network density reduced as constraint conditions, and whether connecting edges exist between adjacent matrix element values and brain region nodes is judged, so that a brain function network topology is constructed;

and 3, step 3: and (5) extracting the features of the brain function network. The feature extraction method is characterized in that the relation between network features and central neural plasticity indexes is researched by using graph theory, and the topological features of the brain function network with central neural plasticity change under tDCS are described by selecting the feature path length, the clustering coefficient and the betweenness of the directed network;

and 4, step 4: the effect of tDCS on brain function networks was studied. Analyzing the multi-lead EEG signals by adopting a brain function network, drawing undirected graphs of all frequency bands of the EEG according to the correlation matrix obtained in the step 1, and calculating brain function network indexes: connectivity rate and small world characteristics. Comparing the brain function network indexes before and after the tDCS stimulation, thereby revealing the characteristic change of the brain function network connection relation of the premotor, the motor area and the motion perception area of the stimulated hemisphere during the motion activity.

And step three, extracting brain-brain, brain-muscle and muscle-muscle coupling characteristic indexes from the data collected in the step one by adopting a multi-level neuromuscular coupling analysis method based on the causal measurement of the dynamic regression model, and using the extracted characteristic indexes to represent the rehabilitation effect of the motor function. The specific process is as follows:

step 1: the EEG and EMG signals are pre-processed separately. Aiming at an EEG signal, firstly, carrying out band-pass filtering of 0.1-100Hz on the EEG signal, and simultaneously filtering out baseline drift and noise interference of power frequency; then, combining the prior knowledge of the ocular artifacts, and eliminating the artifacts in the EEG signals by adopting a reference independent variable analysis method; and then extracting corresponding frequency band signals in each channel in the EEG signal, such as 1-4 Hz delta waves, 4-8 Hz theta waves and 8-12 Hz alpha waves, according to the states of the study object and the subject in the experiment, wherein the waveforms in different frequency bands correspond to different physiological characteristics. For the EMG signals, firstly, a 50Hz wave trap is adopted to filter power frequency interference in the EMG signals, and then band-pass filtering of 0.1-100Hz is carried out;

step 2: and (3) optimizing and selecting the electroencephalogram electrodes related to the upper limb movement by adopting a typical correlation analysis CCA theory to obtain an optimal electroencephalogram lead set. Because of the dispersion of EEG signals, there is a certain individual difference in EEG signals associated with upper limb movement, and therefore, the EEG signals associated with upper limb movement need to be determined, and the EEG signal lead set with the highest degree of association with upper limb movement is optimally selected. Calculating the correlation degree of an EEG signal and an EMG signal during specific upper limb movement by adopting a CCA method, and selecting an electroencephalogram electrode with higher correlation degree with the specific upper limb movement as an optimal electroencephalogram lead set for subsequent electroencephalogram and electromyogram coupling analysis by taking a CCA coefficient as a reference basis;

and 3, step 3: causal values between multi-lead EEG-EEG, EEG-EMG and EMG-EMG signals are calculated using causal measurements based on a dynamic regression model. In order to embody the change information of the causal value along with the signal frequency, different frequency bands are extracted from a time sequence by adopting a wavelet packet decomposition method and a multivariate empirical mode decomposition method, and then the causal value among multi-lead signals under each frequency band is calculated;

and step four, guiding the stimulation mode and parameters of the tDCS in the next step by researching the correlation between the brain-brain, brain-muscle and muscle-muscle coupling characteristic indexes and the influence rule of the tDCS on the neural plasticity, so as to achieve the plasticity management of the motor cortex rehabilitation.

The method comprises the steps of extracting that a patient is in different rehabilitation stages by adopting a multi-level neuromuscular coupling analysis method, describing progress changes of upper limb movement function recovery by nerve oscillation relations among different areas of the brain, central nerves and muscle movement responses and motor muscle units during training, and then carrying out rehabilitation evaluation by a clinician by using a Fugl-Meyer upper limb movement function evaluation scale. If the rehabilitation effect becomes better after weak stimulation is carried out at the original tDCS stimulation position, the stimulation intensity can be further enhanced; if the rehabilitation effect becomes poor after the original tDCS, the stimulation position can be changed or the original anode stimulation can be changed into cathode stimulation. The rehabilitation evaluation provides effective basis for further tDCS stimulation mode and parameter improvement, and effective guidance and management of nerve remodeling can be realized by adjusting the tDCS intervention mode.

Claims

1. Dyskinesia cortical plasticity management device based on transcranial electrical stimulation brain muscle coupling, its characterized in that: the device comprises a rehabilitation experiment module and a management module; wherein the content of the first and second substances,

the rehabilitation experiment module adopts a transcranial electrical stimulation tDCS device to form an electrical stimulation intervention channel of cerebral cortex, and utilizes a multi-channel EEG signal acquisition function of the tDCS device and a wireless EMG signal acquisition device to respectively form EEG and EMG signal acquisition channels;

the management module is used for executing the following steps:

step 1, receiving data collected by a rehabilitation experiment module;

step 2, extracting brain function network characteristics reflecting the characteristics of neural activity by adopting a brain function network modeling method, and revealing the influence rule of tDCS on the activity of sensory and motor cortex and the plasticity change of central nerves;

step 3, extracting brain-brain, brain-muscle and muscle-muscle coupling characteristic indexes from the data obtained in the step 1 by adopting a multi-level neuromuscular coupling analysis method based on the causal measurement of a dynamic regression model, and evaluating the rehabilitation effect of the exercise function;

step 4, guiding the stimulation mode and parameters of the tDCS in the next step by researching the correlation between the brain-brain, brain-muscle and muscle-muscle coupling characteristic indexes in the step 3 and the influence rule of the tDCS in the step 2 on the sensory motor cortex activity and the central nerve plasticity change, and realizing plasticity management of motor cortex rehabilitation;

the specific process of step 3 is as follows:

3-1, preprocessing EEG and EMG signals respectively: aiming at an EEG signal, firstly, carrying out 0.1-100Hz band-pass filtering on the EEG signal, and simultaneously filtering out baseline drift and noise interference of power frequency; then, combining the prior knowledge of the ocular artifacts, and eliminating the artifacts in the EEG signals by adopting a reference independent variable analysis method; then extracting corresponding frequency band signals in each channel in the EEG signals; for the EMG signals, firstly, a 50Hz trap filter is adopted to filter power frequency interference in the EMG signals, and then 0.1-100Hz band-pass filtering is carried out;

3-2, adopting a typical correlation analysis CCA theory to optimize and select the electroencephalogram electrodes related to the upper limb movement to obtain an optimal electroencephalogram lead set: selecting the correlation degree of an EEG signal and an EMG signal during upper limb movement by adopting a CCA method, and selecting an electroencephalogram electrode with higher correlation degree with specific upper limb movement as an optimal electroencephalogram lead set for subsequent electroencephalogram and electromyogram coupling analysis by taking a CCA coefficient as a reference basis;

3-3. calculating causal values between the multi-lead EEG-EEG, EEG-EMG and EMG-EMG signals using causal measurements based on a dynamic regression model: extracting different frequency bands from a time sequence by adopting a wavelet decomposition method and a multivariate empirical mode decomposition method, and then calculating causal values among multi-lead signals under each frequency band to be used as brain-brain, brain-muscle and muscle-muscle coupling characteristic indexes;

2. The device for managing cortical plasticity of movement disorders based on transcranial electrical stimulation brain muscle coupling according to claim 1, wherein: the specific process of step 2 is as follows:

2-1, establishing a brain function network correlation matrix: defining the measured area of each lead of the EEG as a node of the network, wherein the electrical activity of the EEG is a plurality of time sequences, establishing the connection relation between every two EEG signals of each channel, and calculating the intensity value of the EEG signals, thereby establishing a correlation matrix;

2-2, establishing an adjacency matrix and brain function network topology: on the basis of the correlation matrix, converting the correlation matrix into a sparse adjacent matrix through threshold processing, and judging whether a connecting edge exists between an element value of the adjacent matrix and a brain area node, thereby constructing a brain function network topology;

2-3, extracting the characteristics of the brain function network: researching the relation between the network characteristics and the central neural plasticity indexes by using graph theory, and selecting the characteristic path length, the clustering coefficient and the betweenness of the directed network to describe the brain function network characteristics of the central neural plasticity change under tDCS;

2-4. correlation analysis of motor cortex brain function network characteristics under tDCS: analyzing the multi-lead EEG signals by adopting a brain function network, drawing undirected graphs of all frequency bands of the EEG according to the correlation matrix obtained by 2-1, and calculating the index of the brain function network: and comparing the brain function network indexes before and after the tDCS stimulation so as to reveal the characteristic change of the brain function network connection relation of the premotor, the motor area and the motion perception area of the stimulated hemisphere during the motion activity.