A review on - EEG signals by motor imagery based brain computer interface

Abstract

Motor Imagery- Brain Computer Interface (MI-BCI) is known to be a recent blooming technique since it acts as non- muscular channel that helps for disabled people for communication. A exact and trustable MI-BCI system needs the extraction of enlightening and accurate features because it is a communication link among the wired brain and an external device. A brain computer interface, or BCI, plays a crucial role in human beings, particularly disabled persons. BCI is one in which various brain functions are analyzed or controlled with the help of computers (controllers). Along with-it EEG signals could be utilized to study the brain activity and also to diagnosis of diseases like neurological disorder, Parkinson's disease. EEG signals can be processed and the features can be extracted for the dimensionality reduction with the help of various techniques which includes PCA, Fast Fourier Transformation, Wavelet Transformation. These extracted features can be classified using classifiers such as SVM, K - Nearest neighbors, Neural network and few other methods. Now a days, the role of deep learning (DL) technique has had aowing impact on Motor imaginary-EEG-based Brain Computer Interface and data set has a vital role in all applications. Here in this paper, we have reviewed the recent papers that applied Physio net dataset to analyze and evaluate electroencephalogram (EEG) signals for various applications which contains 64 EEG channels and 109 subjects executing diverse tasks.