Brain Computer Interface-Based Signal Processing Techniques for Feature Extraction and Classification of Motor Imagery Using EEG: A Literature Review

A communication path for people having severe neural disorders is provided by Brain Computer Interaction. The Brain–Computer Interface in an electroencephalogram is an important and challenging one for managing non-stationary EEG signals. EEG signals are more vulnerable to noise and artifacts. The Motor Imagery-based Brain–Computer Interface is used as a communication channel for people with neural disorders who have no muscular activity. For a well-established and accurate BCI system, two important steps have been used in MI-BCI, such as feature extraction and feature classification. Spectral methods and spatial methods are used for the feature extraction methods. The classifiers translate the features into the device commands. Linear Discriminant Analysis is the most widely used classification algorithm. So far, Support Vector Machine has been used as a classification method. In recent studies, Deep Neural Networks and Convolutional Neural Networks have been used. In this study, the feature extraction approaches as well as the signal classification methods for the motor imagery brain computer interface are thoroughly reviewed and presented.