Boosting Cognitive Focus via Attention Types Detection using Brain-Computer Interfaces: A Pilot Study

Abstract

This study leverages Brain-Computer Interfaces (BCIs) and electroencephalography (EEG) to enhance cognitive focus in adolescents (12–17 years) by classifying effective (task-oriented) and ineffective (distracted) attention states. Addressing declining attention spans in Generation Alpha/Z, we integrate augmented reality (AR) environments with personality-adaptive machine learning models. Sixteen participants performed cognitive tasks while EEG data was captured via a 16-channel BrainAccess MIDI headset. Signal preprocessing (filtering, ICA- independent component analysis, CSP -common spatial patterns) tied with data augmentation improved dataset robustness by 40%. Results demonstrated a 57% concentration increase in AR versus VR (where participants performed identical tasks in a non-adaptive virtual environment) with personality-tailored models boosting classification accuracy by 10%. High-performing classifiers (e.g., Deep Neural Networks, XGBoost) achieved 87% accuracy, underscoring BCIs’ potential for personalized cognitive interventions in education and therapy.