Abstract
Purpose :
Choriocapillaris (CC) imaging yields great potential to gain essential understanding of pathophysiological processes in various retinal diseases. However, due to anatomical properties and overlaying structures, imaging artifacts can greatly alter results. To our knowledge, only the superficial retinal plexus was analyzed as a potential disruptive factor, whereas the subjacent layer of the CC, the choroid, has not yet been investigated. This study addresses the impact of choroidal vessels on CC flow deficit analyses.
Methods :
In this cross-sectional study, macular 6 x 6 mm swept source optical coherence tomography angiography scans centered on the fovea were obtained, using a Zeiss PLEX Elite 9000 device and analyzed within four 1 x 1 mm squares at the corners of each image. The CC was defined as a 20µm thick slab starting at the lower border of the Bruch’s membrane. Structural information of the slab was used for signal compensation. Images were binarized and analyzed with ImageJ. The presence of drusen and choroidal vessels was identified by two retinal experts, if no consensus was reached, a third expert was consulted. Two-factorial ANOVAs were used to identify possible effects of choroidal visibility and sector localization (nasal and temporal) on flow deficit percentage (FD%) in compensated and uncompensated images.
Results :
Seventy-seven eyes of 46 patients with intermediate age-related macular degeneration (AMD) were included in the analysis. After exclusion of sectors containing drusen, 112 sectors with and 138 sectors without visible choroidal vessels were analyzed. Sixty-two (80%) of all patients showed visible choroidal vessels in at least one sector.
The ANOVAs showed a significant lower FD% in sectors containing visible choroidal vessels in compensated as well as uncompensated images (p < .001). A significant main effect of sector localization, with lower FD% in the nasal sector, was only detected in compensated (p < .001) but not uncompensated images (p = .053). This finding might be attributed to vessels of the superficial plexus in these sectors. There was no significant interaction in both analyses.
Conclusions :
The visibility of choroidal vessels significantly affects the quantification of CC flow deficits in SS-OCTA images. Researchers should pay close attention to this type of artifact and consider removing affected areas from further analysis.
This is a 2020 ARVO Annual Meeting abstract.