Examination of Optimal Mental Task for Brain-computer Interface using Functional Near-infrared Spectroscopy

Abstract

Brain-computer interface (BCI) is a new technology that enables communication for patients in totally locked-in state using brain activities. In this research, we examined optimal mental task for near-infrared spectroscopy-based BCI from among six mental tasks. SVM was used for learning, and the classification accuracy for each tasks through 8-fold cross validation was calculated. As a result, it was suggested that the music imagery had high discrimination accuracy and was a practical task.