Abstract
Purpose :
The study of retinal vessel density is crucial in evaluating disease severity and progression especially in retinal diseases such as diabetic retinopathy. Unfortunately, not all optical coherence tomography angiography (OCTA) machines have the capability of measuring vessel density. Hence, our study aims to describe a novel technique of evaluating retinal vessel density using OCTA images.
Methods :
In a prospective cohort study of 14 volunteers (7 healthy and 7 pathological cases), 3mm x 3mm OCTA scans were performed on both eyes using the AngioVue OCTA system. OCTA scans of the superficial retinal vasculature were exported and assessed using ImageJ (version 1.49, W. S. Rasband, Maryland, USA). The images were binarized using a pre-selected auto-thresholding method. Vessel density of each image was obtained by calculating the percentage of area occupied by retinal vessels. Calculated vessel density values were then compared to the density values generated by the AngioVue OCTA system.
Results :
The mean vessel density was 41.96 (range 38.37 – 55.21) on the AngioVue OCTA system and 48.72 (range 28.60 – 53.18) as measured by our method (p=0.242). The difference between the 2 methods ranged from 3.57% to 25.47% with a mean of 6.71%. There was no significant difference between the 2 methods when comparing between the eyes of our male and female volunteers (mean difference of 17.77% and 10.07% respectively; p=0.087). Neither was there a significant difference between the 2 methods when comparing normal and pathological eyes (15.35% and 12.50% respectively; p=0.549).
Conclusions :
Our method is comparable to the automated density measurement as provided by the AngioVue OCTA system with an error of 6.71%. It is hence reliable in serving as a viable method of measuring retinal vessel density in both normal and pathological eyes.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.