Multi-acquisition averaging optical coherence tomography angiography for diabetic retinopathy

Abstract

Purpose : To study the impact of multi acquisition averaging on qualitative and quantitative evaluation of diabetic retinopathy (DR) with optical coherence tomography angiography (OCT-A).

Methods : Ten eyes of 10 patients with DR were included in this study. OCT-A images were acquired from an OCT clinical prototype with real-time speckle variance processing software. The OCT system used a 1060nm swept source (Axsun Inc) with 200 kHz A-scan rate and a 105 nm wavelength sweeping range. High performance computing was used for real time visualization of the retinal microvasculature during acquisition and to produce quantitative analysis in a clinically useful time frame. For each eye, 10 OCT-A volumes centered at the foveal avascular zone (FAZ) were serially acquired over X seconds. The patients were asked to fixate, but were able to blink as required. The serially acquired images were divided into strips automatically, identifying blinks and eye motion. Image discontinuities and warping were corrected by strip-wise affine registration and subsequent local similarity-based non-rigid registration of the strips.

Results : Representative images of the full thickness, superficial and deep retinal capillary layers before and after motion corrected averaging are shown in Fig 1 (A) and Fig 1 (B), respectively. The non-averaged images have a poor signal to noise ratio resulting in spiky, non-smooth vessel contour and poor visualization of the deep capillary plexus where the signal is weaker. White linear artifacts and image lateral displacement caused by micro-saccades results in vessel discontinuity and overall poor identification of the capillary network. The higher quality of averaged images can be appreciated by the presence of vessels with smooth contour, absence of motion artifacts and better identification of the deep capillary plexus.

Conclusions : High quality visualization of the retinal microvasculature, in particular the DCP, may improve our understanding of the onset and development of retinal vascular diseases such as diabetic retinopathy.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.