Signal Processing And Classification Methods for Brain-Computer Interfaces


Ashkan Yazdani


Prof. T. Ebrahimi




Electroencephalogram based Brain Computer Interface (BCI) is a system that empowers the users to communicate with their environments and also control external devices not through their neuromuscular pathways but only with their brain’s electrical activities. This could be of significant importance when the users suffer from diseases which lead to major Paralysis’s or when users have injuries in their Central Nervous Systems (CNS) and consequently have no control over their limbs. The application’s spectrum of BCI is not only limited to this but also encompasses other interesting purposes such as rehabilitation and therapy.

A typical BCI system usually comprises of signal acquisition and feature extraction and finally pattern classification of the recorded signal. In this Thesis different feature extraction, feature selection and classification techniques will be used to develop such system with the best performance possible. The goal of this PhD thesis is to extend further the research previously fulfilled in the area of BCI and EEG signal processing and to develop a new state of the art BCI system.