This project builds a compact, reliable Oculomotor Function Index from public eye-tracking data. The plan centers on GazeBase, a large, multi-task, longitudinal dataset (322 participants; 12,334 recordings; tasks spanning fixation, saccades, reading, free viewing, and more). Features to extract include parameters of the saccade main sequence and saccade latencies, fixation stability and microsaccade rate, smooth-pursuit gain, and reading scanpath statistics. Reliability is quantified with the intraclass correlation coefficient and with subject-held-out generalization. The scientific aim is a normative score with confidence intervals that remains stable within individuals yet varies across individuals.
References
Griffith, Henry, et al. "GazeBase, a large-scale, multi-stimulus, longitudinal eye movement dataset." Scientific Data 8.1 (2021): 184.
Virtual-Eye Project: https://www.simula.no/research/projects/virtual-eye-learning-human-eye-scanpaths-optimal-autonomous-search