EEG-based Brain-Machine Interface (BMI) using machine learning to control upper limb robotic arm for rehabilitation

Abstract

Recently, research on brain computer interface (BCI) for rehabilitation of stroke patients has become popular. Rehabilitation reflecting patient’s motor intention is more effective for motor recovery compared to passive robot-assisted or therapist based rehabilitation. If the command sent to the robot reflects patient’s motor intention, muscle movement and brain’s motor circuit would become more synchronized, resulting in enhanced neural plasticity. Thus, this study examines the possibility of decoding motor intention during and without movement along with classification of various activities of daily life (ADL) for rehabilitation.