Brain Biomarkers for Assessing Practice of Robotic Arm Reaching Movements Using a Head-Controlled Interface.

Limited efforts have examined the cognitive-motor processes as individuals learn to operate upper-limb assistive devices that improve interactions in their environment (e.g., prostheses, head-controlled devices). Prior work mainly focused on performance without examining cerebral cortical dynamics via brain biomarkers to assess the level of practice during the learning of such assistive devices. Specifically, it was suggested that EEG biomarkers like low- and high-beta spectral power are related to learning as they assess memory formation and attentional mechanisms. Therefore, this work examines how sensorimotor performance, and low-/high-beta spectral power are influenced as individuals without disabilities practice reaching movements with a simulated robotic effector executed via a head-controlled interface. Results revealed faster and straighter reaching movements. Additionally, as individuals progressed from early to late practice, the movement planning stage revealed an elevation of frontal, central and parietal low-beta power and an attenuation of high-beta power in the temporal region. These spectral modulations may reflect an internal model memory encoding process of the novel sensorimotor mapping imposed by this interface throughout practice which ultimately enables improved reaching performance. This work could inform patients' cognitive-motor processes when learning to control assistive systems and provide biomarkers for monitoring practice during rehabilitation.