Quantitative automated circumpapillary microvascular density measurements: a new angioOCT-based methodology

Purpose To develop a new methodology to detect glaucoma damage based on circumpapillary microvascular density (cpmVD) as measured by optical coherence tomography angiography (angioOCT). Methods AngioOCT scans from a random eye of 40 healthy subjects (aged 63 +/- 14 years) and 82 glaucoma patients (aged 66 +/- 9 years with an average visual field loss of -7.8 +/- 6.5 dB) were used to develop a new angioOCT evaluation methodology. Optic disc-centred 3 x 3 mm images were collected (Angioplex (R), Zeiss Cirrus 5000 HD-OCT). An annular area with an inner radius of 1.16 mm and outer radius of 1.44 mm was chosen as the region of interest (ROI), with cpmVD calculated through lower envelope modulation avoiding the influence of the major retinal vessels. Additionally, the cpmVD and the microvascular density (mVD), stratified by sectors, were compared with the respective retinal nerve fibre layer (RNFL) thickness. Results A significantly lower cpmVD was observed in the glaucoma group, for all visual field sectors (p < 0.001), when comparing to the healthy group. The inferior and superior mVD sectors showed the largest differences between groups. Moreover, glaucoma was linked to a reduction of cpmVD variation when compared to the healthy group. Lastly, a comparison between two healthy subgroups showed that the method is not biased by the presence of slight movement artefacts, optic disc dimensions and vessel crowding (p > 0.05). Conclusions The presented open-source methodology provides a robust quantitative analysis of the 360 degrees mVD. It shows that cpmVD, and mVD sectors measured by angioOCT, can be used in everyday glaucoma practice.