Towards a brain computer interface driven exoskeleton for upper extremity rehabilitation

Abstract

Stroke impairs individuals to perform activities of daily living. Intense rehabilitation programs offer hope for recovery, but are labor intensive and costly. Robotic rehabilitation technology plays a key role to solve such a problem. Current robotic systems along with brain computer interface (BCI) allow patients to participate in rehabilitation exercises, which require their own mental inputs. Studies have shown such active rehabilitation exercise can induce neuroplasticity and help towards recovery. However, even though BCI-driven robotic systems do exist, they are large complex systems and expensive to set up. These drawbacks limit a wide distribution of these technologies. Currently, the BCI robotic systems only used in large hospitals or research settings, not community level facilities. To facilitate the accessibility of stroke patients to such technologies, we propose a novel BCI-driven exoskeleton rehabilitation system. The exoskeleton has four degrees of freedom (DOF) for assisting the movement of the upper extremities. It is integrated with an affordable and wireless EEG headset for enabling the patients to control the movement of the exoskeleton with their brain activity. The developed exoskeleton is portable and easy to set up. A sequential control scheme is proposed to allow the user to control one movement at a time. An experiment was designed to assess if a healthy individual was able to control the movement of the exoskeleton correctly under a predefined sequence. One volunteer participated in the exploratory study and the volunteer was able to correctly control the exoskeleton in each step.