Development of an EEG Controlled Wheelchair Using Color Stimuli: A Machine Learning Based Approach

Brain-computer interface (BCI) has extensively been used for rehabilitation purposes. Being in the research phase, brainwave based wheelchair controlled systems suffer from several limitations, e.g., lack of focus on mental activity, complexity neural behavior different conditions, and lower accuracy. sensitive to color stimuli, EEG signal changes promises a better detection. Utilizing Electroencephalogram (EEG due methodology by brainwaves presented this study. Red, Green, Blue (primary colors) Yellow (secondary color) were chosen as stimuli utilized 2 × window four-direction command, namely left right, forward stop. Alpha, beta, delta theta rhythms analyzed, time frequency domain features extracted find most influential rhythm accurate classification model. Four classifiers, namely, K- Nearest Neighbor (KNN), Support Vector Machine (SVM), Random Forest Classifier (RFC) Artificial Neural Networks (ANN) trained tested assessing performance each rhythm, with five-fold cross-validation. measures, i.e. sensitivity, specificity, accuracy area under receiver operating characteristic curve examine wholescale performance. The results suggested that Beta performs best apart all control. While comparing ANN-based classifier shows 82.5%, which is higher than three other classifiers.