A Novel EEG Classification Technique Based on Particle Swarm Optimization for Hand and Finger Movements

Abstract

Electroencephalogram (EEG) has gained much attention from researchers recently. EEG classification has many applications such as: classifying brain disorders, helping paralyzed people to control a machine by their own imagery mental tasks and controlling a robot or a remote system with both imagery and actually mental tasks. This paper aims to classify arm and finger movements acquired through EEG signals. The EEG signals have been transformed to frequency domain using discrete wavelet transform (DWT) as a feature extractor. These extracted features are then feed into a novel particle swarm classifier to classify the different movements of arm and fingers. The experimental results showed that this new algorithm gives accuracy of 95% with minimum time delay, which is an essential requirement for all biomedical applications. This research is considered the first step towards implementing an automatic system (surgical robot) that can be used in Telesurgery.