Dataset Information

Optical Coherence Tomography Structural Abnormality Detection in Glaucoma Using Topographically Correspondent Rim and Retinal Nerve Fiber Layer Criteria.

ABSTRACT:

Purpose

This study evaluated the ability of topographically correspondent (TC) minimum rim width (MRW) and peripapillary retinal nerve fiber layer thickness (pRNFLT) criteria to detect optical coherence tomography (OCT) structural abnormality in glaucoma (GL) and glaucoma suspect (GLS) eyes.

Design

Retrospective cross-sectional study.

Methods

A total of 196 GL eyes, 150 GLS eyes, and 303 heathy eyes underwent pRNFL and 24 radial optic nerve head OCT imaging and manual correction of the internal limiting membrane, Bruch's membrane opening (BMO), and outer pRNFL segmentations. MRW and pRNFLT were quantified in 6 Garway-Heath or 12 30-degree (clock-hour) sectors. OCT abnormality for each parameter was defined to be less than the 5th percentile of the healthy eye distribution. OCT abnormality for individual eyes was defined using global, sectoral, and combined parameter criteria that achieved ≥95% specificity in the healthy eyes. TC combination criteria required the sectoral location of MRW and pRNFLT abnormality to be topographically aligned and included comMR (a previously reported TC combination consisting of MRW and pRNFLT parameter: [MRW + pRNFLT × (average MRW healthy eyes/average pRNFLT healthy eyes) MRW].

Results

TC sectoral criteria (1 Garway-Heath MRW + corresponding Garway-Heath RNFLT), (one 30-degree MRW + any 1 corresponding or adjacent 30-degree pRNFLT), 30-degree and Garway-Heath comMR-TI and global comMR were the best performing criteria, demonstrating (96%-99% specificity), 86%-91% sensitivity for GL, 80%-84% sensitivity for early GL (MD ≥ -4.0 dB) and 93%-96% sensitivity for moderate-to-advanced GL (MD < -4.0 dB).

Conclusions

Clinically intuitive TC MRW and pRNFLT combination criteria identified the sectoral location of OCT abnormality in GL eyes with high diagnostic precision.

SUBMITTER: Yang H

PROVIDER: S-EPMC7214190 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:PRéCIS:: There are errors in automated segmentation of the retinal nerve fiber layer (RNFL) in glaucoma suspects or patients with mild glaucoma that appear to persist over time; however, automated segmentation has greater repeatability than manual segmentation.PurposeTo identify whether optical coherence tomography (OCT) segmentation errors in RNFL thickness measurements persist longitudinally.MethodsThis was a cohort study. We used spectral domain OCT (Spectralis) to measure RNFL thickness in a 6-degree peripapillary circle, and exported the native "automated segmentation only" results. In addition, we exported RNFL thickness results after "manual refinement" to correct errors in the automated segmentation, and used the differences in these measurements as "error" in segmentation. We used Bland-Altman plots and linear regression to determine the magnitude, location, and repeatability of RNFL thickness error in all twelve 30-degree sectors and compared the error at baseline to follow-up time points at 6 months, 2 years, 3 years, and 4 years.ResultsWe included 406 eyes from 213 participants. The 95% confidence interval for errors at baseline was -6.5 to +13.2 μm. The correlation between the baseline error and the errors in the follow-up time periods were high (r>0.5, P<0.001 for all). Automated segmentation had a smaller SD of residuals from the longitudinal trend line when compared to manual refinement (1.56 vs. 1.80 μm, P<0.001), and a higher ability (P=0.009) to monitor progression using an analysis of a longitudinal signal-to-noise ratio.ConclusionsErrors in automated segmentation remain relatively stable, and baseline error is highly likely to persist in the same direction and magnitude in subsequent time periods. However, automated segmentation (without manual refinement) is more repeatable and may be more sensitive to glaucomatous progression. Future segmentation algorithms could exploit these findings to improve automated segmentation in the future.