Abstract
Purpose :
While numerous studies have focused on characterizing the choroid in central serous chorioretinopathy (CSCR), visualizing the retinal pigment epithelium (RPE) remains challenging. This prospective study aimed at using Adaptive Optics-Transscleral Flood Illumination (AO-TFI) images to assess RPE cell morphology in areas that appear healthy in CSCR eyes and compared them to those of healthy eyes.
Methods :
The study included 12 CSCR eyes (2 active, 10 resolved) from 14 patients (age 42.3 ± 5.9 years) and 33 eyes from 19 healthy volunteers (age 36.7 ± 12.6 years). For each eye, macular 5°x5° AO-TFI images were obtained using the prototype Cellularis retinal camera. Quantitative image analysis included two approaches: tessellation, identifying cell centers and boundaries with the Voronoi algorithm, and segmentation, detecting hypo-reflective areas within RPE cells. Morphological analyses were conducted using both methods, and the results were statistically compared between healthy and CSCR populations.
Results :
: RPE morphological features were quantified in 54 AO-TFI images from CSCR eyes and 149 from healthy eyes. Significant differences were observed in the morphological heterogeneity of RPE cells in CSCR eyes compared to healthy eyes. The tessellation analysis showed that the standard deviation (SD) of the number of neighbors was significantly higher in CSCR RPE (P=0.001), indicating greater dispersion in cell organization. The segmentation analysis indicated smaller average circularity (P=0.008) and solidity factors (P=0.0054), along with increased morphological heterogeneity compared to healthy RPE cells.
Conclusions :
This study reveals morphological changes in retinal regions that appear healthy in CSCR, suggesting stress on the RPE cell cytoskeleton. Prospective longitudinal studies and correlation with choroid-specific imaging techniques are needed to confirm the role of epitheliopathy in CSCR pathology. RPE imaging using AO-TFI holds promise for improving our understanding of CSCR pathogenesis and early diagnosis.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.