Evaluation of sensorimotor rhythms to control a wheelchair

Brain activity produces electrical signals that can be detected in a non-invasive or invasive way, and BCI (Brain-Computer Interface) systems can translate these signals into commands, which allow communication with devices without involvement of peripheral nerves and muscles. This paper shows the development of a non-invasive BCI system using brain signals from somatosensory cortex. The acquisition of signals was performed using one electrode cap in accordance with 10-20 system and using 3-channel EEG in order to command a motorized wheelchair without involvement of peripheral nerves and muscles. In the experiment with the wheelchair, 65.7% hit rates (average) were obtained for two directions, respectively.