Astigmatism effect and its correction in optical coherence tomography angiography imaging

Purpose: To study the influence of astigmatism on the qualitative and quantitative analysis of optical coherence tomography angiography images. Methods: We included patients with an astigmatism of >0.5D and those without or with an astigmatism <0.5D. In patients without astigmatism, we induced a cilyndrical error of -1 and -2D by placing a lens in front of the camera head of the Spectralis. In patients with astigmatism, we corrected it using the same method. In both groups, we performed two macular optical coherence tomography angiography (OCTA) scans, with and without astigmatism, in a random order. We analyzed the differences in terms of vessel density (VD), image quality and presence of artifacts in images acquired with and without astigmatism in the superficial vascular plexus (SVP) and deep vascular complex (DVC). We also performed a linear regression to investigate the interaction between diopters and VD. Results: We included 110 eyes of 110 patients (20 in the induced astigmatism group and 90 eyes with >0.5D). We found that eyes with astigmatism of –2D showed a greater reduction in VD values, worse image quality and more defocus and attenuation artifacts. We report a VD reduction of 0.012-0.02 per diopter of astigmatism. The axis of did not show any significance in our cohort. Conclusions: Astigmatism of -2D significantly affects VD values and image quality mainly in the SVP and should be corrected when performing OCTA analysis. A cylindrical lens coupled to the camera head of the Spectralis device is a valid method to improve OCTA images by correcting astigmatism in these cases.