Improved Black Widow Optimization Assisted Channel Selection for Motor Imagery BCI:A Non-Parametric Analysis

Many years back, BCIs are utilized for helping persons with disabilities in the brain. In most of these applications, BCI is developed to accommodate over activities that people can no longer complete on their own. For instance, brain-computer interfaces (BCIs) might be utilized for communication as well as for controlling wheelchairs, robotic arms, and other medical equipment as well as orthoses & prostheses. Some other strategy is to employ a BCI for rehabilitation in which the user can recover more normal or normal function in the motor rather than replacing the motor function directly through control of a BCI. One popular method for helping stroke victims regain their motor function is motor imagery (MI). Therefore, this article aims to describe a new channel selection-based MI BCIs strategy. In this work, the Gauss Mutated Black Widow Optimization Algorithm (GMBWOA) selects the suited channel in this case. For the congruent selection of optimal channels and solution samples, an objective model is created. Bayesian Linear Discriminate Analysis (BLDA) is used to classify the selected channels and features that were recovered from the original signal using Common Spatial Pattern (CSP). The final resultant from BLDA provides the classified Motor Imagery (MI) EEG signal. Additionally, the results of the proposed system are examined with the existing approaches by employing a variety of metrics.