Abstract
Purpose :
SD-OCT segmentation algorithms currently do not perform well in segmenting intraretinal layers in eyes with Stargardt disease. These inaccuracies can be managed by tedious manual correction. We compared selective B-scan segmentation strategies in generating mean retinal layer thickness and preserved area data from SD-OCT scans in patients with Stargardt disease (STGD1).
Methods :
13 eyes from 13 STGD1 patients were randomly selected from the ongoing Natural History of the Progression of Atrophy Secondary to Stargardt Disease (ProgStar) studies. For each eye, a 49-section 20°x20° high resolution SD-OCT scan was collected. The space between each B-scan was approximately 120-125 µm. All 49 B-scans were segmented manually to quantify retinal pigment epithelium (RPE), photoreceptor outer segments (OS) and inner segments (IS), outer nuclear complex (ONC), and inner retina (InR). For each volume scan, mean thickness (MT) and total area (TA) were generated with three different B-scan selection strategies, using: all 49 B-scans; 1 of every 2 B-scans; and an “adaptive” method, containing a subset of (or all) 49 B-scans that the grader deemed as significantly different from adjacent B-scans. The Mann-Whitney-Wilcoxon test was used to test for significant (α = 0.05) differences.
Results :
No statistically significant (p < 0.05) differences were detected between the three strategies (Table 1). For the adaptive strategy, the average # of segmented B-scans was 35 (SD=10; range=22-49).
Conclusions :
These data suggest that a selective segmentation strategy may be a viable alternative to exhaustive, comprehensive manual segmentation in Stargardt disease. These results may have implications for increasing the efficiency of SD-OCT grading strategies in clinical trials for Stargardt disease and other related macular degenerative disorders.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.