Classification of Gamified Go/NoGo Tasks using Single-Channel EEG Device and Machine Learning

Patients who are unable to speak face significant challenges in expressing needs and preferences, leading to isolation and a diminished quality of life. Brain-Computer Interface (BCI) can offer a promising solution, enabling simple and accessible communication through the analysis of brain signals. This paper investigates the feasibility of using a single-channel commercial electroencephalogram (EEG) device for binary communication classification (Go/NoGo tasks). EEG data are collected from 11 healthy participants using a single-channel consumer grade EEG device while they engaged in a decision-making game involving two stimuli: ‘angel’ and ‘demon’. After processing the signal data, 470 features are extracted for the classification. Recursive feature elimination with cross-validation is performed to ensure the most effective result. Explainable Artificial Intelligence (XAI) is used to understand the feature selection process. Among different machine learning models, the Decision Tree performs the best with 82.75% overall accuracy and a recall score of 82%. This research highlights the effectiveness of low-cost, single-channel EEG devices in no-verbal communication tasks, making advanced BCI technology more accessible for applications for healthcare sector to improve quality of patients life.