Abstract
Purpose :
High quality of OCT-angiography (OCTA) images is a prerequisite for the subsequent quantification of retinal microvasculature. We hypothesize that glaucoma affected the prevalence and types of artifacts in OCTA images. The objective of this study was to evaluate the prevalence, distribution, and related factor with en face images derived from swept source OCTA.
Methods :
In this retrospective, observational analysis, en face OCTA images in macular region from glaucoma and healthy eyes were obtained. Two graders independently assessed the 3×3 mm2 OCTA images of superficial vascular plexus. The influence of age, sex, diagnosis, visual acuity, spherical equivalent, axial length, intraocular pressure and signal strength(SS) on artifacts were evaluated by logistic regression.
Results :
OCTA images of 343 eyes of 183 subjects involving 100 patients and 83 healthy participants were included for evaluation. At least one type of artifact was present in the images from 303 eyes (88.34%). Totally 11 kinds of artifacts including projection (343 eyes,100%), motion artifact (258 eyes,75.22%), blurred images (85 eyes,24.78%), out of center (73 eyes,21.28%), blink artifact (9 eyes,2.62%), vignetting artifact (7 eyes, 2.04%), error segmentation (4 eyes,1.17%), stretching artifact (3 eyes,0.87%), unmasking (3 eyes, 0.87%), vessel doubling (2 eyes,0.58%), and out of window (1 eyes,0.29%) were found. The diagnosis of glaucoma (P＝0.012) and lower signal strength(P＜0.001) were significantly associated with the presence of blurred images, after adjusting other factors.
Conclusions :
OCTA imaging artifacts are common and diverse. OCTA image in glaucoma patients have more possibility of blurred image, which requires manual refocusing in clinical practice. Strengthening the recognition of OCTA artifacts and performing rigorous image quality management before images analysis are essential.
This is a 2020 ARVO Annual Meeting abstract.