Abstract
Purpose :
To compare intrasession repeatability of the peripapillary vessel parameters of optical coherence tomography angiography (OCTA) in glaucoma and non-glaucoma subjects.
Methods :
An observational, cross-sectional study collected three 6x6mm optic nerve head (ONH) OCTA scans (Cirrus HD-OCT 5000) of glaucoma and non-glaucoma subjects. En face images of the radial peripapillary capillary (RPC) layer were created using a commercially provided automatic segmentation software. A custom image quality grading protocol excluded eyes with media opacities, segmentation errors, or motion artifacts. Subjects with non-glaucomatous optic neuropathy, diabetic retinopathy, and other retinal pathologies were excluded. The commercially provided quantification software calculated perfusion density (PD) and flux index (FI); custom quantification software calculated vessel area density (VAD) and flux. Repeatability was measured by intraclass correlation coefficient (ICC), within-subject standard deviation (Sw), coefficient of repeatability (CRw), and coefficient of variation (CVw).
Results :
One eye was chosen at random from each subject. Of 193 subjects, 159 subjects had at least two good quality intrasession images. Of these, 88 met the inclusion and exclusion criteria (29 non-glaucoma, 59 glaucoma). The Sw [%], and ICC of the commercially available segmentation algorithms in non-glaucoma eyes for PD were 0.006, and 0.939, respectively, and for FI were 0.006, and 0.954. The Sw [%], and ICC in glaucoma eyes for PD were 0.007, and 0.965, and for FI were 0.007, and 0.969. The Sw [%], and ICC of the custom segmentation software in non-glaucoma eyes for VAD were 0.019, and 0.858, respectively, and for flux were 0.008, and 0.864. The Sw [%], and ICC in glaucoma eyes for VAD were 0.012, and 0.972, and for flux were 0.004, and 0.976.
Conclusions :
OCTA parameters were highly repeatable during the same visit. Furthermore, repeatability measurements were similar for both commercially provided and custom quantification software. ICCs tended to be higher among glaucoma subjects compared to non-glaucoma subjects, which may be explained by the larger between-subject variability in the glaucoma group.
This is a 2020 ARVO Annual Meeting abstract.