Neuro-ocular Dynamics in Stress Detection: An Integrated Analysis of EEG and EOG Modalities
摘要
This research investigates the dynamic interrelation between neurological and ocular markers in response to stress, employing an integrated analysis of electroencephalogram (EEG) and electrooculogram (EOG) modalities. Utilizing 28-channel EEG gel-based electrodes of brain product, we extracted frequency band information, Hjorth parameter, spectral entropy, and statistical parameters, while 2-channel vertical EOG data provided insights into blink rate and the number of blinks. The late fusion of these modalities was implemented for machine learning algorithms, complemented by unimodal prediction strategies. This study reveals findings that are intricate stress-related patterns at the neuro-ocular interface, offering a comprehensive exploration of stress dynamics. Late fusion techniques demonstrated enhanced predictive capabilities, enriching our understanding of stress responses. By elucidating the interconnected roles of the brain and eyes, this study contributes novel perspectives to stress assessment methodologies, potentially shaping future research in the field.