Facilitating independent living of individuals with neurological disorders

Abstract

Facilitating independent living of individuals with neurological disorders, who have upper extremity (UE) impairment, is a compelling goal for our society. The degree of impairment could be reduced by using Electroencephalography (EEG) controlled assistive devices. The successful implementation of EEG controlled devices strongly relies on the capability of properly determining individuals’ actions. Therefore a preliminary study was conducted to evaluate the performance of a classification scheme based on extracting time domain features of the EEG signal. Specifically, the feature vectors were built by extracting root mean square (RMS), autoregressive (AR) coefficients and waveform length from EEG signals. Multi-class support vector machine (SVM) was used as a classifier and an acceptable classification error rate (less than 14%) on average was obtained. It was observed that the classification of three right-arm movements, namely rest, grasp and elbow flexion, was possible in principle.