Online Asynchronous MI-BCI System Experimental Paradigm Design

The Brain-Computer Interface(BCI)is a system that reads the EEG signals of human brain and converts them into computer instructions,and outputs them to an external device to perform direct control of the external devices without the involvement of the brain nerves and normal muscles.BCI technology helps patients with neurological or muscle disorder to achieve wheelchair driving,text input,voice output and other functions.Experimental paradigm is one of the main research directions of the BCI.A well-designed experimental paradigm controls the experiment process by giving hints to users,and collect the EEG signals containing sortable characters gernerated by users,then The classified calculation results are converted into instructions and output to external device,and realize the function of BCI.Based on the basic experimental paradigm of Motor Imagery(MI),an online asynchronous experimental paradigm was designed,avoiding the screen dependency,and improving degrees of freedom and accuracy.In this work,an off-line training experimental paradigm was firstly designed based on motion imaging.The EEG acquisition system containing digital electroencephalograph,electrode cap,electrodes and electrode wires was bulit.Based on the basic experimental paradigm of MI,the idle state was introduced into the experimental paradigm as a new degree of freedom.The feasibility of this improvement was proved by experiment.In order to reduce user fatigue,the duration of the offline training experimental paradigm was shortened by Optimize the time for motor imagery and rest,without the reduce of MI-BCI accuracy.To further improve the degrees of freedom,Steady State Evoked Potential(SSVEP)was introdued into online asynchronous experimental paradigm.The performance of SSVEP was improved by optimizing the channel combination and time window length for algorithm.The practicality of the online asynchronous experiment paradigm was tested by designing a small game that controls the small squares to touch the target rectangle,and scores of the game was used to evaluate the feasibility of the experimental paradigm.Judgment algorithm in experimental paradigm was improved,and classified MI signal was converted into output instruction to control the small square to move to the left,right,and upward.One degree of freedom was added by SSVEP,which controls the step back instructions for the small squares.The experiment result indicates good performance of the online asynchronous experiment paradigm designed in this work.Furthermore,the design idea of combining different experimental paradigm will provide great help in the practical application of BCI.