Comparison of Wearable Video-based Eye Tracking and EOG for Oculomotor Activity Detection in Specific Research Tasks

Oculomotor activity measurement is crucial for understanding visual perception and cognitive processes in neuroscience. This study conducts a comparative analysis of wearable video-based eye tracking (Pupil Core) and electrooculography (EOG) in detecting eye movements during a neurophysiological task based on Raven’s Advanced Progressive Matrices (APM). Fourteen calibration sessions were performed using five calibration points, and data synchronization was achieved through the Lab Streaming Layer protocol. Both eye tracking methods were calibrated using optimal affine transformation matrices to minimize discrepancies between fixation points and calibration markers. The calibration error averaged 66.4 pixels for the Pupil Core system and 64.73 pixels for EOG data, indicating comparable accuracy between the two methods. Despite EOG’s susceptibility to artifacts from muscle contractions and wire movements, its performance suggests it is a viable alternative to video-based systems, particularly in environments with challenging lighting conditions or when long-term monitoring is required.