EEG-based interaction with virtual worlds: A self-paced three class Brain-Computer Interface

In this paper, we present self-paced Brain-Computer Interface (BCI) based interaction with a computer game. The BCI is able to detect three different motor imagery related brain patterns (imagination of left hand, right hand, foot or tongue movements) in the ongoing brain activity by using three bipolar electroencephalogram (EEG) channels only. To ensure that brain activity was used for control, on-line reduction and detection of eye movements and other muscle artifacts were used. The task of the game was to navigate through a virtual environment and collect scattered coins within a limited time. Results of three subjects are reported. All three subjects managed to control the BCI by using the self-paced operation mode and collect coins. Two out of the three succeeded to collect all of them.