Mental task motor imagery classifications for noninvasive brain computer interface

A Brain computer interface (BCI) has introduced new scope and created a new period for developers and researchers giving alternative communication channels for paralysed peoples. Motor imagery refers to where EEG signals that being obtained while the subject is imagining or performing a motor response. This work is to examine this area from Machine Learning and exploit the Emotiv System as a cost-effective, noninvasive and also a portable EEG measurement device. The experiment was carried out based on Emotiv control panel focusing on cognitive commands such as (forward, backward, left and right). The data were preprocessed to remove the artifact as well as the noise by using EEGlab toolbox. Wavelet transforms namely Daubechies and symlets were used for feature extraction. The Multilayer perception (MLP), Simple logistic and Bagging were utilized to classify the mental tasks motor imagery. The performance of classifications was tested and satisfactory results were obtained with the accuracy rate 80.4% using the Simple logistic classifier.