EEG-Based Analysis of Breakfast's Impact on Cognitive Performance Using Earth Mover's Divergence

Cognitive performance, particularly focus and attention, is critical for academic success, and lifestyle factors such as breakfast consumption significantly influence these functions. However, existing studies often rely on subjective measures, leaving a gap in objective, quantitative assessments. To address this, we propose a novel framework, Optimal Representative Selection via Earth Mover's Divergence (ORS-EMD), which leverages Earth Mover's Divergence (EMD) to objectively quantify the impact of breakfast on cognitive performance using electroencephalogram (EEG) signals. The ORS-EMD framework computes pairwise dissimilarities between EEG segments using EMD, identifies optimal representative segments for each group, and classifies new EEG segments as healthy (breakfast-consumed) or unhealthy (breakfast-skipped) based on a divergence thresholding rule. Our results demonstrate that breakfast consumption significantly enhances cognitive performance, as evidenced by faster task completion times (6.1 ±0.5 minutes for the breakfast group vs. 8.3 ±0.7 minutes for the no-breakfast group, p < 0.001) and distinct neural activity patterns. The ORS-EMD framework achieved an accuracy of 92.5%, outperforming traditional methods such as time-domain features (85.3%) and frequency-domain features (88.7%). This study provides a quantitative, evidence-based framework for understanding the impact of breakfast on cognitive performance, highlighting the potential of EEG-based analysis in real-world educational and healthcare settings.