Research on Motor Imagery EEG Classification Method based on Region of Interest and Electrophysiological Source Imaging Technology

The spatial resolution of EEG signals is cross compared with that of MEG signals, which makes the brain-computer interface (BCI) system using this technology unable to meet online applications. Electrophysiological source imaging (ESI) technology can make up for the shortcomings of EEG signals through the mapping process and achieve higher classification accuracy. However, the mapping process will lead to an increase in the amount of calculation, which will affect the realization of the online BCI system. This study innovatively proposed a ROI division combined with scalp electrode position, and combined it with ESI technology to improve the shortcomings of slow response time of BCI system based on ESI technology. The experimental analysis shows that the method of combining ESI technology and ROI division can shorten the response time and improve the analysis performance, which is more suitable for online BCI system.