Noninvasive control of a robotic arm in multiple dimensions using scalp electroencephalogram

• Published in: 2013 6th International IEEE/EMBS Conference on Neural Engineering (NER) pp. 45 - 47
• Main Authors: Baxter, Bryan S., Decker, Andrew, Bin He
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.11.2013
• Subjects: Computers; Educational institutions; Electroencephalography; Neural engineering; Robot sensing systems; Time-frequency analysis
• ISSN: 1948-3546
• DOI: 10.1109/NER.2013.6695867

Robotic limbs can enable individuals with impaired mobility and motor system output to interact with their physical environment. Many of these individuals do not have the motor control necessary to utilize their hands to control an assistive device. We hypothesize both healthy and post-stroke individuals can modulate sensorimotor rhythms with motor imagery to control of a human-size robotic arm and hand to complete multi-step grasping tasks using noninvasive EEG signals. Two healthy and one post-stroke subject controlled a robotic arm in two-dimensions to move to a target, grab the target block, and move it to a final location using motor imagery recorded through scalp EEG. Subjects demonstrate successful control of a robotic arm in 2D performance of a multi-step task in a time-limited manner.