Poster Session: Deviation mapping for foveal cone mosaic topography

Deviation mapping is commonly used across retinal imaging modalities. Here we compiled data from two labs (UC Berkley[1] & MCW) to create an AOSLO-specific deviation mapping tool for measures of the foveal cone mosaic. Foveal cones were identified for 87 normative regions of interest (ROIs) (26M, 61F; 13-67 yrs, median=26 yrs) and for 5 pathological ROIs (2 Bornholm Eye Disease, 3 Albinism; 1M, 4F; 16-50 yrs, median=42 yrs). ROIs were cropped and resized to a common scale for comparison. Density and nearest neighbor distance (NND) maps were generated for each ROI, and the cone density centroid[2] (CDC) was determined for each map. Normative maps were aligned using these CDC locations, and average and standard deviation (SD) maps were created for both density and NND. Pathology maps were compared to these normative composite maps. At the CDC, average (SD) density was 1.79E+5 (2.55E+4) cones/mm^2 and average (SD) NND was 2.08 (0.16) µm. For pathological ROIs, the percentage of pixels within 1 SD of the normative data was comparable for density and NND except in two individuals where density was more deviant than NND (consistent with mosaic irregularity and/or random cone loss). Deviation mapping applied to foveal AOSLO data can be used to assess the normality of individual foveal ROIs. Comparing deviation maps across different metrics may provide valuable insight into the underlying properties of the cone mosaic in various retinal pathologies. 1) PMID:31348002 2) PMID:34343479.