Abstract
Purpose :
To quantify the vascular density of extended areas of the retina, images of sufficient linear resolution must be used. Current wide-field scans, while useful to visualize large areas in a single image, may not have sufficient resolution to allow accurate quantification of vessel and perfusion densities. In order to overcome this limitation, and to do vascular quantification on larger areas, we demonstrate the use of higher resolution montaged images in the quantification algorithm.
Methods :
For each of 5 subjects, a series of 3 OCTA scans having 6 mm x 6 mm FOV was acquired on a PLEX© Elite 9000 SS-OCT System (ZEISS, Dublin, CA). These scans were montaged into a single wide-field image using Zeiss proprietary software from the Advanced Retina Imaging (ARI) portal.
Since the standard ETDRS grid is too small to cover the large areas in wide-field images, an extension of this grid was used to capture vascular density in areas lateral to the standard grid.
The composite superficial retinal layer image was enhanced and processed to allow calculation of perfusion and vessel linear density. These densities were then averaged over each of the areas in the extended grid.
Five normal subjects were measured three times each to assess the repeatability of these measurements. The square root of the variance due to repetition, calculated by means of ANOVA, was divided by the global mean to calculate the intra-individual relative standard deviation (RSD) of the measurements by grid position.
Results :
Fig.1 shows the modified ETDRS grid used in this study. A table of the output of the quantification algorithm for the single test subject, along with the RSD due to measurement variation, is shown in Table 1. RSD’s of less than 10% were obtained for all non-foveal grid positions.
Conclusions :
The use of montaged OCTA scans allows vascular density quantification in wider fields of view than is possible in single scans, because the resolution of montaged images can be significantly greater than that of a single wide-field scan.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.