The Performance Of A Novel P300 Brain-Computer Interface Paradigm With Electrical And Vibration Modes

Abstract

A novel tactile P300 paradigm was proposed for Brain-Computer Interface(BCI), in which only two types of stimuli was used to distinguish different targets. We also adapted an algorithm for intention recognition, which used spatial information to distinguish different stimulation sites, and frequency information to identify attended-target stimuli and disturbances. Our novel paradigm was verified on both vibration and electrical stimuli modes, and achieved an average classification accuracy of 95.21% for vibration stimuli mode and 94.88% for electrical stimuli mode, respectively. Furthermore, we evaluated performances of brain's functional connectivity by the correlation coefficient, and preliminarily explored the similarities and differences between vibration stimuli mode and electrical stimuli mode.