Abstract
The topography of S-cones in the human retina is vital to understand short-wavelength sampling of visual space. In humans S-cones have been reported as randomly arranged within 2° eccentricity and semi-regular more peripherally. A model describing how S-cone regularity varies across the retina is yet to be formulated. Here we describe such a model, dependent on 2 parameters - the average distance between neighboring cones and the proportion of S-cones - that is sufficient to explain S-cone regularity across the central retina. Cones were classified using AO-OCT optoretinography in ROIs distributed across the 4 cardinal meridians in 2 subjects (12 ROIs each) between 1.3 - 12.9°eccentricity. The radius of the S-exclusion zone, the area surrounding S-cones where other S-cones are significantly unlikely to appear, was found to be about twice the average distance between neighboring cones in 19/24 mosaics. We found that the measured regularity of S-cone mosaics increases linearly with the increasing proportion of S cones with eccentricity. Using the average distance between neighboring cones and proportion of S-cones per ROI as variables, we created a model to simulate S-cone mosaics that agree well with the observed topography. These results benefit our understanding of the foundational patterns underpinning spectral topography, and the ability to accurately simulate S-cone topography in computational models of early vision.
Funding: Funding: NIH grants 5T32EY007031-42, U01EY032055, EY029710, P30EY001730. Burroughs Wellcome Fund Careers at the Scientific Interface. DOD Air Force Office of Scientific Research FA9550-21-1-0230. University of Washington eScience Institute Incubator Program.