Abstract
Purpose :
To evaluate the influence of astigmatism errors in the qualitative and quantitative analysis of optical coherence tomography angiography (OCT-A) images.
Methods :
Patients undergoing a comprehensive ophthalmic evaluation including OCT-A scans were included in the study. The OCT-A images were acquired using Spectralis HRA-OCT2 (Heidelberg Engineering, Heidelberg, Germany). In the same visit, 10°x10° High-Resolution OCT-A images were acquired with and without astigmatism correction using the follow-up tool. Qualitative analysis (no changes, subtle changes, substantial changes) was performed by 2 independent graders (RDM, RGP). Quantitative analysis included skeletonized vessel density (sVD) of the superficial, intermediate and deep capillary plexus (SCP, ICP, DCP respectively) and the choriocapillaris (CC). Statistical analysis was performed using SPSS.
Results :
Five hundred eyes of 264 patients were included in the present study. Median astigmatism error was -0.75 diopters [range: 0.25-5.50]. Qualitative analysis was significantly influenced by astigmatism error greater than 2.50 diopters (p<0.05). In the uncorrected scans, artefactual increase of vessel diameter of the greater retinal vessels was observed within the SCP, associated with a decrease in the visualization of the small capillaries in the SCP, ICP, and DCP. Values of sVD of the SCP, ICP, DCP and CC were significantly influenced by astigmatism errors (p<0.001).
Conclusions :
OCT-A images provide valuable depth-resolved information on the vascular layers of the retina and choroid. However, different factors influence the quantitative analysis of OCT-A images, limiting our ability to extract conclusive data from this relatively novel modality. Our results suggest that correction of significant astigmatism errors may be required in studies including quantitative OCT-A analysis.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.