Artifact rates for 2D retinal nerve fiber layer thickness versus 3D neuroretinal rim thickness using spectral domain optical coherence tomography

Abstract

Purpose : The minimum distance band (MDB) is a 3D neuroretinal rim measurement derived from high-density spectral domain optical coherence tomography (OCT) volume scans of the optic nerve using customized research software. Although the MDB has equal or better diagnostic capability for glaucoma compared to the traditional 2D peripapillary RNFL (pRNFL) thickness parameter, the artifact rate of this newer 3D parameter is still unknown. We performed a cross-sectional, observational clinical study to compare the rate of clinically significant artifacts per eye for 2D pRNFL thickness versus 3D MDB thickness.

Methods : Only one eye per patient was used for analysis in 120 glaucoma patients and 114 normal patients. A total of 15 artifact types were recorded for 2D and 3D OCT scans, and artifact rates were calculated per B-scan and per eye. In 3D volume scans, artifacts were recorded for only the B-scans which intersected the optic nerve and which were used for final neuroretinal rim calculations. Global MDB thickness values were calculated twice, once before and then once after manual deletion of B-scans with artifacts and subsequent automated interpolation. A clinically significant artifact is defined as one that requires a repeat scan in order to obtain an accurate measurement. Pre- and post-interpolation measurements were considered equivalent if they differed by less than normal inter-test variability.

Results : Among glaucomatous eyes, artifacts were present in 71.7% of 2D pRNFL thickness scans (86 of 120) and 20.9% of B-scans (1423 of 6820) in 3D MDB volume scans. Mean global MDB thickness was similar before and after interpolation in both the glaucoma group (169.3 ± 50.7 μm versus 176.1 ± 50.7 μm, respectively, p-value=0.30) and the normal group (303.8 ± 44.5 μm versus 309.4 ± 43.0 μm, respectively, p-value=0.33). Three-dimensional scans were less likely to have clinically significant artifacts, with 15.8% (19 of 120) of glaucomatous eyes and 13.2% (15 of 114) of normal eyes with artifacts compared to 2D pRNFL scans with 71.7% (86 of 120) of glaucomatous eyes and 38.6% (44 of 114) of normal eyes with artifacts (p-value<0.0001 for both groups).

Conclusions : Compared to the traditional 2D pRNFL thickness scans, 3D optic nerve volume scans less frequently require manual correction and repeat scanning to obtain accurate measurements.

This is a 2020 Imaging in the Eye Conference abstract.