4-Class Motor Imagery Classification For Post Stroke Rehabilitation Using Brain-Computer Interface

Abstract

Brain-Computer Interface (BCI) is a mechanism that helps in the control/communication of one’s environment through the brain signals obtained directly from the brain via an EEG signal acquisition unit. A BCI incorporating Motor Imagery for post-stroke rehabilitation of upper limbs and knee in fully disabled patients is designed. It helps in restoring some of the activities of the daily living. It aids post-stroke sufferers to carry out functionalities like movement of right and left hands, right and left knee, grabbing things, etc. Till now, the post-stroke rehabilitation is possible in partially impaired patients. The therapy was given to the damaged cortical area in the brain directly by the therapist moving the hand of the patient. This is not suitable as it requires the presence of therapist, can be done only in a hospital leading to high cost, muscle fatigue, slow recovery of the damaged area. In this paper, a survey of various BCIs is done and an optimum design is devised. Also the overall accuracy of the system is calculated to be approximately 91%.