Hybrid-BCI: Classification of auditory and visual related potentials

The brain computer interface (BCI) is a technology that utilizes neurophysiological signals recorded from brain to control external machines or computers, and has become widespread in the last decade due to technical and mechanical developments. A P300-based BCI, often called P300 speller, is one of the most successful paradigm, which has shown advantages in terms of high accuracy and short training time. However, the existing P300-based BCI employs single type of external stimuli, such as visual stimuli, which limits their application domains. In this paper, we propose a hybrid-BCI system based on multiple modality of P300 evoked by simultaneous auditory and visual stimuli. The experimental results show the significant difference in ERPs between visual stimuli and multiple types of stimuli. The classification results demonstrate the effectiveness of our new BCI paradigm, which outperforms the visual P300 in terms of higher accuracy.