An on-line BCI system for hand movement control using real-time recurrent probabilistic neural network

This paper presents a new EEG-based Brain-Computer Interface (BCI) for on-line controlling the hand movement in a virtual reality environment. The goal of this research is to develop an interaction technique that will allow the BCI to be effective in real-world scenarios for hand grasp control. For this purpose, two classifiers are designed. The first classifier which is based on the imagination of right-hand movement is for controlling the hand grasping, holding and opening. The second classifier, which is based on the imagination of left-hand movement is designed for error correction and activating the first classifier. One important issue in developing an on-line BCI is the robust and accurate classification of EEG signal which is characterized with a time-varying distribution. In this work, we present a real-time recurrent probabilistic neural network for classifying the EEG signals. The results show that the subjects were able to achieve an accuracy more than 80% during the first session of experiment without off-line training and 73%-91% during the last session using single-trial classification with no adaptation.