Related Experiment Video
Updated: Aug 17, 2026
Evaluation of Capillary and Other Vessel Contribution to Macular Perfusion Density Measured with Optical Coherence Tomography Angiography
Published on: February 18, 2022
Impact of Optical Coherence Tomography Scan Pattern and Chorio-Scleral Interface Delineation on Choroidal Vascularity
Kryshell Yu Qi Wong1, Rachel Ka Man Chun1,2,3, David Alonso-Caneiro4,5
1Centre for Myopia Research, School of Optometry, The Hong Kong Polytechnic University, Kowloon, Hong Kong.
Purpose:
To compare radial and horizontal raster acquisition patterns using spectral-domain optical coherence tomography (OCT) for assessing choroidal vascularity index (CVI) in healthy eyes and to evaluate how variability in chorio-scleral interface (CSI) delineation influences CVI measurements.
Methods:
OCT images of 25 eyes from 25 healthy adults were acquired using 12-line radial and 31-line horizontal raster scans with enhanced depth imaging. CVI was measured manually using three different posterior boundaries-defined as the upper (vascular), middle (vascular-stromal) and lower (stromal) CSI boundaries. Agreement between CSI boundaries and scan patterns was assessed using Bland-Altman analysis. Topographical CVI variations across 6-mm Early Treatment of Diabetic Retinopathy Study (ETDRS) subfields, quadrants and rings were evaluated using two-way repeated-measures ANOVA.
Results:
Participants were aged 21-31 years (72% female; mean spherical equivalent -2.97 ± 2.02 D). Mean CVI differences ranged from 1.43 to 2.38% between the upper and middle CSI boundaries and from 1.23 to 2.04% between the middle and lower CSI boundaries. Comparisons between the upper and lower CSI boundaries resulted in larger CVI differences, ranging from 2.90 to 4.42%. The coefficient of variation for CVI was low (<2%) between radial and raster scan patterns, and Bland-Altman analysis showed overall good agreement. The 31-line raster scan measured 0.51% higher CVI in the central foveal region, whereas the 12-line radial scan yielded 0.64-1.10% higher CVI in the perifoveal region. Significant mean differences were observed only in the outer temporal subfield (-1.10%, p = 0.005) and outer ring (-0.77%, p = 0.02). No significant scan-pattern effect was found, but CVI varied significantly across the ETDRS grid, with the lowest values in the outer nasal subfield, nasal quadrant and outer ring.
Conclusions:
Radial scans provide CVI measurements comparable to raster scans, with faster acquisition, supporting practical use in children. Standardising CSI delineation is crucial for consistent CVI comparisons across studies.