EEG Acquisition and Motor Imagery Classification for Robotic Control

Abstract

The adoption of brain-computer interfaces (BCIs) has significantly increased in various application domains, particularly in the field of controlling robotic systems through motor imagery. The article contributes in two primary ways: 1) validating the effectiveness of using a minimally invasive electroencephalography (EEG) device combined with machine learning techniques to control fundamental movements in a robotic setting, and 2) demonstrating these findings practically through the construction of a robotic vehicle. In this vehicle, tasks involving motor imagery align directly with control commands for the vehicle. To validate our approach, we identified four-class and two-class classification tasks. The signals have been acquired from a portable EEG device equipped with eight dry electrodes. We employed sliding window strategies to segment the data, along with feature extraction using the Common Spatial Pattern (CSP) method. Classification modules were implemented based on Support Vector Machine (SVM) and K-Nearest Neighbors (KNN) models. The experimentation involved five participants, each with their own personalized model. While the accuracy of results in the four-class tasks is not notably high, the outcomes in binary classification tasks are promising, boasting an average accuracy of approximately 61%. Results suggest a promising potential for this approach in the realm of robot control, particularly when employing dry-electrode EEG devices.