Dynamics of Human Learning of a Brain-Computer Interface

Brain-machine interfaces represent a novel output mechanism of human brain activity, where a neuronal-based signal is used to directly control an external device (e.g., a computer cursor, robotic arm, etc). In our model, subjects are patients who undergo temporary implant of electrocorticographic (ECoG) arrays directly on the brain surface for clinical mapping of function and epileptic focus. The functional areas have a reliable signal that can be used to drive a computer cursor.  The learning of this occurs, behaviorally, over repeated attempts.  Brain changes outside of the control region of cortex can also be measured with ECoG.  The changes associated with learning reveal experience-dependent changes in activity, in areas that overlap with normal motor learning. Changes during sleep and behavioral transitions further delineate the changes in the brain over the course of improved performance.