A novel grid strategy for correlating focal macular anatomic changes with focal changes in choriocapillaris perfusion

Abstract

Purpose : To determine the relationship between focal anatomic changes in the macula with focal changes in choriocapillaris (CC) perfusion over time, we developed a novel grid strategy for registering these features on swept-source OCT (SS-OCT) angiography scans, but first we needed to determine the test-retest repeatability of our strategy.

Methods : In this retrospective review of eyes enrolled in a prospective SS-OCT imaging study, normal eyes, eyes with intermediate AMD (iAMD), and AMD eyes with hypertransmission defects (hyperTD) were studied. Each of the eyes selected had three sequential, same-day 6mmX6mm SS-OCTA scans. For each scan, hyperpigmentation (hyperPig), calcifying drusen (CaD), hyperTDs were identified and a combined mask of these features was made. Using a previously published algorithm, we applied the masks and calculated the percentages of CC flow deficits (CCFD%) for both 3mm and 5mm fovea-centered circles for each scan; CCFD binary maps were also generated. Next, the three sequential scans, with their masks and maps, were registered using the retinal vessels as landmarks. To account for differences in the masked regions between scans, the registered masks were merged and applied to each registered map. A fovea-centered grid was overlayed onto these final maps and CCFD% measurements were obtained for each of the grid’s 25 boxes and for the total grid region. The 95% minimal detectable change (MDC₉₅) was calculated for the CCFD% measurements.

Results : Ninety eyes of 63 patients (30 normal eyes, 30 eyes with iAMD, and 30 eyes with hyperTDs) were used in this study. Within the 3mm circle, MDC₉₅ for normal, iAMD, and hyperTD eyes were 1.96%, 1.70%, and 1.89%, respectively. Within the 5mm circle, the MDC₉₅ were 1.36%, 0.85%, and 0.97%. For the individual boxes in the grid, the MDC₉₅ were 3.40%, 3.75%, and 4.88%. For the total grid area, the MDC₉₅ were 1.37%, 0.98%, and 1.24%. Taking all eyes into consideration, the MDC₉₅ within the 3mm circle, 5mm circle, individual boxes, and the total grid area were 1.85%, 1.08%, 4.06%, and 1.21%.

Conclusions : By using the minimal detectable change for CCFD% measurements within different regions of interest, we can now determine if a given change in the CCFD% measurements represents a real change that can be correlated with any identified anatomic change of interest using SS-OCT imaging.

This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.