Review of the Literature on Brain Computer Interface (BCI) Feature Extraction Using EEG Signals Ahsan Mumtaz (a*), Iman Elawady (b,a), Ismail Rakip Karas (a)
a) Department of Computer Engineering, Faculty of Engineering, Karabuk University, Karabuk, Turkey
*Samahsan540[at]gmail.com
b) Department of Electrical Engineering, Ecole Nationale Polytechnique d^Oran Maurice Audin, Oran, Algeria
Abstract
Brain-Computer Interface (BCI) systems are presented as a means for disabled individuals incapable of motor response to interface with a computer via brain signals. The purpose of a BCI is to convert brain activity into a digital command for a computer. How to extract characteristics from EEG data with random time-varying fluctuations and classify them as precisely as feasible is a major challenge in contemporary BCI research. Techniques for feature extraction are used to extract the features that indicate a unique property extracted from the pattern of brain signal. Historically, EEG analysis was limited to eye inspection alone. Visual assessment of the signal is very subjective and does not lend itself to standardization or statistical analysis. In order to quantify the information in the brain signal, a number of distinct approaches were developed. There are numerous linear and nonlinear approaches for feature extraction. This paper aims to provide a basic introduction to EEG signals and the BCI technology. The report also contains an overview of the typical techniques used for signal feature extraction.
Keywords: Brain Computer Interface (BCI), EEG Electrocardiography,Signals,signal-to-noise ratio (SNR)
