Abstract
Purpose :
The Deep Capillary Plexus (DCP) is thought to be the region of initial insult in Diabetic Retinopathy (DR). However, single frame OCT-A images of the DCP are often insufficient for robust quantification of this layer. The purpose of this study is to investigate the qualitative and quantitative improvements in OCT-A images of the DCP in DR eyes as a result of serial acquisition and averaging.
Methods :
Ten serially acquired OCT-A 2x2mm images centered at the fovea were acquired from twelve DR eyes (ranging in severity from mild non-proliferative DR to proliferative DR) using a swept-source 1060nm OCT clinical prototype. For each volume, the enface images of the superficial, deep, and summed layers were divided into micro-saccade free strips, which were then registered to each other and averaged. This algorithm was also applied on ten 1x1mm images acquired with our Adaptive Optics OCT-A system and compared to one 3x3mm scan acquired the same day with a Zeiss PLEX Elite OCT-A, the latter of which features hardware-based motion tracking. All imaging was performed with REB approval.
Results :
A comparison of the single frame and averaged OCT-A images is shown in Fig. 1. The average SNR of the superficial plexus angiograms increased from 18.81 ± 2.50 dB in the single frame images to 24.23 ± 3.32 dB in the averaged images. For the DCP the average SNR increased from 17.32 ± 1.51 dB in the single frame images to 23.91 ± 2.84 dB in the averaged images.Fig. 2 illustrates the single frame and averaged AO-OCT-A images in comparison with the corresponding region within the PLEX Elite OCT-A image. After averaging, regions containing microaneurysms gain enhanced contrast to the background vasculature and are more readily identifiable.
Conclusions :
Multi-acquisition and averaging of OCT-A and AO-OCT-A images results in qualitative and quantitative improvement in images of DR eyes, especially for the DCP. These methods make quantification of retinal vascular features more reliable, thereby facilitating investigation of their roles in the early pathophysiology of diabetic retinopathy.
This abstract was presented at the 2019 ARVO Imaging in the Eye Conference, held in Vancouver, Canada, April 26-27, 2019.