Research On Hybrid Brain Computer Interface Based On Neuralfeedback

To help patients with motor dysfunction improve their ability to live independently and improve their quality of life,Brain-Computer Interface,which is a kind of technology that directly controls external electromechanical devices by EEG signals,has developed rapidly and has a good research prospect in medical rehabilitation and intelligent control.At present,the research on BCI is mainly based on single mode EEG signal control system,but it has some shortcomings in practical application,such as low signal recognition rate and low information transmission rate.Aiming at the above problems,a hybrid BCI system combining the motor imagination paradigm and the steady-state visual evoked potential paradigm was proposed in this paper.Aiming at the problem of low signal recognition rate in the study of monomodal brain-computer interface based on motor imagination,an online neurofeedback training platform was built to improve the signal recognition rate of the subjects.By the parallel fusion of the trained motor imagination EEG signal and the steady-state visual evoked potential EEG signal,compared with the single type of EEG signal control system,it not only increased the number of recognition tasks and also improved the recognition accuracy of EEG signal,and finally completed the brain-controlled Dobot arm writing experiment.The specific research work of this paper is as follows:(1)Based on motion imagine single mode under the state of the design of the brain-machine interface experimental paradigm,and aiming at the problem of motion imagine EEG recognition rate is low,this article takes sports participants imagine brain electrical signal spectrum in turbulence in the alpha and beta rhythm changes as physiological indexes,through the analysis of power spectral density method to build a display window of the EEG changes in energy,as a sports imagine online neural feedback training system.In the online experiment,the recognition accuracy of electroencephalogram signals in the motor imagination was improved by about 11%after the subjects received the Neurofeedback training.The online experiment results showed that the Neurofeedback training could enhance the subjects' motor imagination ability and improve the signal recognition rate.(2)in the steady-state visual evoked potential experimental paradigm of research process,considering the stimulation of interface design,and according to thefrequency characteristic of EEG signals induced obviously this characteristics,using wavelet decomposition and Fourier transform on the preprocessing and feature extraction respectively,and finally using the processing method of typical correlation analysis on the characteristics of classification,five subjects participated in the BCI based on steady-state visual evoked potential single-mode condition online experiment,finally have completed the design task in advance and obtain the average accuracy of 91.95%.(3)A BCI paradigm was designed for the fusion of motion imagination and steady-state visual evoked potential,in which two types of motion imagination tasks were used to control the lifting or descending of the robot arm,and four types of steady-state vep tasks were used to control the movement of the robot arm in four directions: up,down,left and right.A total of 8 subjects participated in the brain-controlled robotic arm writing experiment.According to the experimental results,it can be concluded that this fusion paradigm has a good classification accuracy and can effectively control the Dobot arm to complete the writing task.It also indicates that the hybrid brain-computer interface system based on neural feedback proposed in this paper has certain application potential and prospect.