Experimental Setup of Computer Assisted Stroke Rehabilitation

Abstract

Stroke is the second leading cause of death in Sri Lanka. Physiotherapy is a major component in the rehabilitation of stroke victims. Functional Electrical Stimulation (FES) is often used as a component of this therapy. Brain-computer interface (BCI) is a system that can interpret the user’s intent and translate it into an executable action. There has been evidence of BCI being used in stroke therapy although the results are inconclusive. While conventional physiotherapy is focused on the passive limb movement, BCI-controlled FES, which utilizes the intent to move the limb to control the device, actively involves the brain in rehabilitation. Innovative rehabilitation techniques are necessary, and such need to be evaluated in the Sri Lankan stroke population, to study its effectiveness. Thus, a clinical study is very appropriate. This paper covers aspects of the experimental setup of the clinical trial at the Teaching Hospital, Karapitiya (THK). The main setup changes were in the equipment domain and the experimental paradigm. In the case of FES, a relay circuit was designed to split the FES signal which later was simplified after the pilot subject. In the case of Electromyography (EMG), a new design was proposed to replace the EMG data acquisition system. In the case of an Electroencephalogram (EEG), an alternate method was needed as the EEG currently at THK was unsuitable. The experiment protocol was modified to accommodate participant requests. The BCI system was developed with the power of different EEG frequency bands and using a support vector machine (SVM) and random forest (RF) algorithms. In future, this exploratory study plans to quantify the neuroplastic changes associated with the recovery and study the improvement (if any) of having the brain involved in rehabilitation. Thus, a clinical trial of this nature is necessary and will bring valuable information to the scientific community.