EEG-channel selection based on difference analysis of mental activities

In this paper, we present a selection method of electroencephalography (EEG) channels for a brain-computer interface (BCI) experiment used to control walking behaviors of a humanoid robot via human mental activities. The goal of this study is to decrease the data dimension of the sampled neural signals via selecting the suitable EEG channels. We used coefficient of determination as an index to evaluate the difference of EEG channels. As an example, we analyzed the data of thinking three different walking behaviors: turning right, turning left and walking forward based on the difference of mental activities, and the result is that a part of EEG channels could be selected to control the robot. We further investigated the relation among six selected channels for the same thinking behavior, and found the similarity between the right and left scalps as well as the difference among the three channels on each half scalp.