Binocular rivalry and optokinetic nystagmus as a screening tool for amblyopia
Poster Presentation 43.417: Monday, May 18, 2026, 8:30 am – 12:30 pm, Pavilion
Session: Eye Movements: Clinical
Schedule of Events | Search Abstracts | Symposia | Talk Sessions | Poster Sessions
Adrien Chopin1 (), Preeti Verghese1, Dennis Levi2; 1Smith-Kettlewell Eye Research Institute, San Francisco, 2Herbert Wertheim School of Optometry and Vision Science, University of California, Berkeley
Background: Amblyopia is the leading cause of monocular vision loss, yet current screening methods lack the scalability required for mass screening in preschool children. Binocular rivalry is a perceptual alternation that occurs when incompatible stimuli are presented to the two eyes. We have recently demonstrated that static binocular rivalry dynamics distinguish amblyopic patients from controls. However, this method relies on subjective reports, which are often unreliable in preschool children. Here, we propose a scalable diagnostic tool leveraging motion rivalry dynamics and Optokinetic Nystagmus (OKN) as an objective biomarker of amblyopia. Methods: Adult participants (7 amblyopes, 7 controls) viewed vertical gratings drifting in opposite directions, which elicited motion binocular rivalry and OKN that follows the dominant percept. Participants reported subjective percepts via keypresses while we simultaneously extracted objective percepts from eye-tracked slow-phase OKN velocities. The same participants reported percepts during static rivalry to replicate previous findings. Feasibility for pediatric populations was assessed using a motion rivalry task adapted for young children. Results: First, we replicated the perfect separation of patients and controls using subjective reports in static rivalry, as we previously reported. Second, we confirmed that motion rivalry could achieve the same result. We then compared OKN-derived rivalry metrics, such as alternation rates or eye dominance, to those derived from manual reports, and found high correlations. Crucially, we achieved perfect patient-control separation when based solely on OKN-derived metrics, and using only a single minute of data (F1-score: 100%). Furthermore, a 6-year-old child yielded a high proportion of usable eye-tracking data (77%) in the adapted task, confirming feasibility for pediatric use. Conclusion: Motion rivalry dynamics provide a highly sensitive signature of amblyopia. By utilizing OKN as a proxy for perception, this method bypasses the need for subjective responses, paving the way for a rapid and automated screening tool suitable for preschool children.