Studies in feature selection technique using immune system principles and its application in brain computer interface

Abstract

A Brain Computer Interface (BCI) is a hardware and software communications system that uses cerebral activity to control computers or external devices. BCI helps provide communication capabilities to severely disabled people who are totally paralyzed or locked inand#8223; by neurological neuromuscular disorders. A BCI system works by recording the brain signals and applying machine learning algorithms to classify the brain signals and performing a computer controlled action. ElectroCorticoGraphy (ECoG) is gaining attention as a recording technique for use in BCI as it is better suited for basic neuroscience research and resulting translational opportunities compared to signals acquired from the scalp ElectroEncephaloGraphic (EEG). In a generic BCI framework, the signal acquisition stage captures the brain signals and may also perform noise reduction and artifact processing. The preprocessing stage prepares the signals in a suitable form for further processing. The feature extraction stage identifies discriminative information in the brain signals that have been recorded. Once measured, the signal is mapped onto a vector containing effective and discriminant features from the observed signals. The extraction of this information is a very challenging task. Brain signals are mixed with other signals coming from a finite set of brain activities that overlap in both time and space. One of the main challenge faced in BCIdesign is high dimensional feature vectors, to improve the classification of the signals, relevant features using feature selection techniques is required. newline