Time-frequency features differentiate direction of finger movement in cued and self-paced tasks

With the intended goal of non-invasive control of prosthetic movement in quadriplegic patients, time-frequency distributions of scalp-recorded event-related potentials (ERPs) of movement of digits on the right hand of healthy subjects were obtained using continuous electroencephalography (EEG) and analyzed to differentiate between motor tasks. Robust, statistically significant differences were observed in the time-frequency features which may be useful in non-invasive identification of the intended direction during cued and self-paced pronation and supination of the right index finger and thumb.