A comparative study and analysis of recent EEG-based brain-computer interface approaches and their applications.

Brain-Computer Interface (BCI) is an innovative tool that interprets brain instructions and enables humans for using their mental instructions to operate computers, robots, and other external equipment. People with motor disabilities may now utilize external devices much more easily because to recent advancements in BCI technology. This study analyzes newly developed models and provides a comprehensive literature overview of current developments for the field of completely EEG-based BCI systems. In this paper, the basic concepts of EEG-based BCIs are presented, and the key constituents of the system, such as electrophysiological control signals, feature extraction, classification, and performance measurements in different BCI applications, are discussed. The paper considers the brain's functional connectivity, which is studied through relations among different brain areas as connectivity patterns that reflect information exchange between them at rest and during certain tasks. A comparison among various deep learning models for EEG-Based BCI with subject dependent mode is also presented. The result shows that ATCNet model gives highest accuracy with 86.46% and the lowest accuracy was given by EEG-TCNet with 50.69.