Abstract
Purpose :
To develop an OCT system that generates high-quality, high-resolution 75-degree field OCTA images. These images could help detect peripheral vascular pathology critical in the management of diabetic retinopathy (DR).
Methods :
We developed a high-speed OCT system using a 400 kHz swept source laser (AXSUN) with 1060 nm central wavelength and 100 nm bandwidth. A wide-field sample arm acquired 75-degree horizontal field of view (FOV) images. We also developed a real-time OCT/OCTA data acquisition and tracking system that generates real-time OCT and OCTA images using a highly efficient split-spectrum amplitude-decorrelation angiography (SSADA) algorithm implemented on a graphics processing unit. The system also detects microsaccadic motion, blinking and other artifacts in real-time, and re-scans as needed.
Results :
We scanned 12 DR patients including 8 with non-proliferative DR (NPDR) and 4 with proliferative DR (PDR). The image comprised 1208 A-lines per B-scan with a 38-degree FOV in the fast axis, and 2304 positions with two repeated B-scans within a 75-degree FOV in the slow axis. The sampling density was 10 µm/A-line on both the fast and slow axes. The total scanning time was less than one minute with the tracking software engaged. En face OCTA images were generated using maximum projection within the inner retina region after automatic segmentation was applied. Ten of 12 images were free of blink and large motion artifacts. All images showed minor motion artifacts. Two images showed both blink and large motion artifacts. All DR patients showed vascular abnormalities in both central and peripheral areas, including nonperfusion and microaneurysms (Fig. 1).
Conclusions :
This novel OCTA imaging system can acquire high-resolution, wide-field images noninvasively in DR, significantly improving the ability to detect peripheral vascular abnormalities with the OCTA without the need for montaging.
This is a 2020 ARVO Annual Meeting abstract.