EEG-Based Detection of ADHD: A Comparison of Keyboard and EMOTIV-Controlled Gameplay

This paper describes an experiment that records EEG data of both ADHD and healthy subjects while they play an EEG-controlled game using the EMOTIV, a wireless EEG reader. The study compares EEG data recorded during EMOTIV-controlled gameplay and traditional keyboard-controlled game-play. EEG signals in four frequency bands $(\theta,\ \alpha,\ high\beta,\ low\beta,\ \gamma)$ were analyzed. Two popular machine learning classifiers, Support Vector Machines, and Random Forest Classifiers, were applied to differentiate between ADHD and non-ADHD individuals. The Random Forest Classifier yielded the best accuracy for the dataset.