Repeatability of FAZ Area and Vessel Density Analysis with Projection Artifacts Removal on Optical Coherence Tomography Angiography (OCT-A) in Normal, Glaucoma and Retina Eyes

Abstract

Purpose : To evaluate repeatability of the automated macular superficial plexus vessel density (SPVD), deep plexus vessel density (DPVD), parafoveal 300 µm ring vessel density (FD-300) and foveal avascular zone (FAZ) area parameters of OCT-A scans with AngioAnalytics software for AngioVue® (Optovue, Inc., Fremont, CA), and to estimate the effect of projection artifacts removal (PAR) on DPVD in 3 study groups.

Methods : Up to 3 consecutive fovea-centered 3X3 mm (304X304 A-scans) OCT-A scans per eye were acquired from 15 normal (N), 16 glaucoma (G), and 13 retina (R) subjects (one eye per subject) through IRB-approved protocol. Automatically generated vessel density parameters before and after PAR (Fig.1) were exported and analyzed for distribution and repeatability.

Results : Total of 42, 40 and 24 qualified scans were analyzed for N, G and R groups, following exclusion of scans with motion or segmentation artifacts.
Whole image (WI) vessel density (VD), foveal 1 mm diameter zone (FZ) VD, FD-300 and FAZ area were evaluated for distribution and repeatability, including superficial plexus vessel density (SPVD) and deep plexus vessel density (DPVD) with and without PAR for WI and FZ areas (Fig.1) – Table 1 summarizes the results.
SPVD was significantly higher than DPVD with PAR in N and R groups (p=0.0002 and 0.0007 correspondingly), while no significant difference was detected in G group (p=0.64). In all groups WI DPVD with PAR was significantly lower than without PAR (p<0.0001 for all groups). For FZ no significant difference was found for DPVD with and without PAR in N and G groups (p=0.25 and 0.1 correspondingly), while in R group the DPVD with PAR compared to without PAR was significant (p=0.0087).

Conclusions : AngioAnalytics OCT-A software provides repeatable readings of FAZ area and vessel density parameters. Application of PAR noticeably affects DPVD in N, G and R eyes, resulting in significantly lower DPVD values than without PAR, implicating the necessity of PAR for deep vasculature plexus evaluation.

This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.

 

Fig.1 Enface deep plexus OCT-A (upper), OCT b-scan (middle) with overlaid OCT-A flow signal in red and vessels density image (lower) of the same scan without (left) and with PAR (right).

 

Table 1 Distribution and repeatability of the study parameters