Abstract
Purpose :
To provide comprehensive structural assessments of key cellular characteristics of the photoreceptor (PR)-retinal pigment epithelium (RPE) complex over the entire macula, primarily as a baseline for future disease-related biomarker development.
Methods :
Two healthy volunteers were assessed using the FDA 3.4 MHz Fourier Domain Mode-Locked adaptive optics OCT (FDML-AO-OCT) platform. One eye per volunteer was imaged. For each volunteer, nine 2°×2° field-of-view (500×500 pixels) regions were captured from the fovea to 12° temporal (T) eccentricity, with the AO focus set to the PR-RPE complex. At each location, 30 volumes were collected with a six second interval to enhance contrast by RPE organelle motility induced speckle. The volumes were registered post-hoc and a new generation artificial intelligence (AI) toolset was leveraged to automatically segment the cone PRs for their density and outer segment length (OSL) measures. RPE cells were counted manually using Voronoi analysis and accounting for vessel shadows. Cone PR density, RPE cell density, cone OSL, and cone-to-RPE ratio were analyzed at 1° intervals from the fovea to 12°T.
Results :
In agreement with previous studies, we find that both cone PR density and OSL decrease significantly with distance from the fovea. The OSL drops by approximately 2-fold from 39.2 µm at 1° to 17.1 µm at 12°T. The cone density decreases more sharply, by over half at 4° and by more than 5-fold at 12°T as compared with the cone density at 1°T. The RPE cell density also decreases with an increase in retinal eccentricity across the macula, although with a lower slope than PR cone density. In our volunteers, we found an approximate two-fold decrease in RPE density across the macula from a mean density of 8×105 cells/mm2 at the fovea to 5×105 cells/mm2 at 12°T. Collectively, the cone-to-RPE ratio decreases from 7.1 to 1.7 from 1°T to 12°T.
Conclusions :
AO-OCT imaging combined with AI analysis allows for accurate volumetric characterization of the PR-RPE complex at a single cell level across the entire macula. These metrics can be used as biomarkers for degenerative diseases that attack the PR-RPE complex once baseline values are established that account for the variability across multiple human subject phenotypes.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.