BCI-based neurorehabilitation and prediction system for stroke patients

Abstract

Many stroke patients are suffered from motor disability and unable to carry out the activity of daily living. The purpose of this project is to develop the prediction and neurorehabitation system for pre-and post-stroke patients. Our system includes twomodules; Human-Computer Interface based stroke early prediction system (iSEP) and the upper limb rehabilitation device controlled by EEG through the use of Virtual Reality (iSRE). ERD/ERS is the phenomenon occurs in the brain that we used to classify motor imagery into left or right side. The virtual hand in the developed flash game will move to the side that the user imagined as well as the device will raise the user's upper limb on the same side of imagery. The results showed that the accuracy of left/right classification algorithm depends on user and training time, approximately 71.25 %.