Martin S. Banks¹ (), Josselin Gautier², Norick R. Bowers³, Austin Roorda⁴; ¹UC Berkeley, ²University of Renne, ³Justus Lubig University, ⁴UC Berkeley

The anatomical region that humans use to fixate and project a target onto the retina is the preferred retinal locus (PRL). Although the PRL is constant across task and time, we do not know whether and how fixations on the PRL, as opposed to nearby locations such as the foveola, affect visual performance. To investigate this, we employed the retinal-image-based eye tracking, and stimulus delivery capabilities of an adaptive-optics scanning-light ophthalmoscope (AOSLO) to measure how stimulus location on the retina -- the where -- and oculomotor timing and behavior -- the when and how -- affect visual acuity. Participants performed a Vernier-acuity task by reporting the direction of offset between two small horizontal bars. Stimuli were flashed very briefly every 2sec in a cadence so participants could adopt a consistent fixation strategy. As the time of presentation approached, participants made microsaccades to position the upcoming target near the PRL, and they started to make fewer microsaccades as stimulus presentation was imminent. When the stimulus fell on or near the PRL, acuity was greater than when it fell on other locations. A few hundred milliseconds after the stimulus was extinguished, drifts and microsaccades reoccurred moving the eye away from where the stimulus had been. Our results show that people exhibit machine-like fixation behavior in space and time to position the upcoming stimulus on a specific retinal location: the PRL. But the results also show that performance is not measurably better or worse in the region of peak cone density, which is roughly 5minarc away.

Acknowledgements: Center for Innovation in Vision & Optics, UC Berkeley