Comparison between the grading of cRORA and iRORA on SD-OCT B-scans with that of persistent hypertransmission defects on en face SS-OCT images

Abstract

Purpose : To compare the gradings of complete and incomplete retinal pigment epithelium and outer retinal atrophy (cRORA and iRORA) detected on spectral-domain OCT (SD-OCT) B-scans with the gradings of persistent hypertransmission defects (hyperTDs) detected on swept-source OCT (SS-OCT) en face images, we performed gradings of same day images from both instruments.

Methods : Patients with late nonexudative AMD were scanned using 6×6 mm macular scans on SD-OCT (Spectralis® Heidelberg, 512x97, ART:9) and SS-OCT (PLEX® Elite 9000, Carl Zeiss Mediec, 500x500 angio pattern) instruments. En face images from the SS-OCT scans were generated from a subRPE slab positioned from 64-400um below Bruch’s membrane. Bright areas, known as choroidal hyperTDs (Fig1B), with a greatest linear dimension (GLD) of ≥ 250µm were designated as persistent choroidal hyperTDs (Fig 1D). On SD-OCT, cRORA and iRORA were graded from horizontal B-scans as regions with outer retinal and RPE attenuation/absence associated with choroidal hyperTDs measuring ≥ 250µm (cRORA) or < 250µm (iRORA). Blinded to SS-OCT scans, the graders examined each SD-OCT B-scan, and the horizontal length of iRORA was labeled with a red line and cRORA with a blue line (Fig1A). The iRORA and cRORA labels were then registered and projected onto the en face subRPE image obtained from SS-OCT scans (Fig 1C, D). Note the grading on SS-OCT was fully automated and fast.

Results : Two examples are shown in Fig 1. cRORA lesions were routinely identified as persistent hyperTDs (Eye 1). However, iRORA lesions could also be identified as persistent hyperTDs on en face imaging because the hyperTDs associated with these lesions measured < 250 µm in the horizontal B-scans, but ≥ 250µm when measured in GLD (both Eyes). Since adjacent B-scans were not considered when grading for iRORA, the full extent of these iRORA lesions were not appreciated when grading horizontal B-scans, leading to ambiguous identification.

Conclusions : The results of this preliminary grading exercise suggest that the full extent of choroidal hyperTDs should be considered in order for lesions to be appropriately designated as iRORA or cRORA. Moreover, grading on SD-OCT is not easily achievable from a B-scan based analysis.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.

 

Fig 1