Area and volume analysis of the superior and inferior retinal nerve fiber layer (rNFL) bundles from SD-OCT thickness maps: description of the methodology

Abstract

Purpose: Inferior and superior thinning of the optic nerve neuroretinal rim is a hallmark of glaucoma. OCT analysis of the peripapillary rNFL thickness is a powerful tool to evaluate early glaucoma. Here we show the usefulness of area and volume analysis of the superior and inferior rNFL bundles for glaucoma evaluation.

Methods: Previously published SD-OCT data from 55 glaucoma and 98 age-matched control subjects was provided by Budenz and colleagues (PMIDs: 22365056 and 21917932). The raw thickness map data (segmented rNFL) from optic disc cube 200x200 Cirrus scans from one randomly selected eye of each subject was obtained from Carl Zeiss Meditec. A low, middle and high threshold value were chosen such that above these values only the inferior and superior bundles were present in the thickness map. The area and volume of the superior (S), inferior (I) and summed superior-inferior (S+I) rNFL bundles were measured using image J software (NIH) for each threshold value. The mean bundle area (mm²) and volume (10^-3 mm³) for the glaucoma and control group were compared using a Student’s t-test.

Results: Using any of the chosen thresholds, the area and volume of all rNFL bundles examined were markedly decreased in the glaucoma subjects (all p < 0.0001). Using the middle threshold, the mean and standard deviation of the areas of the S, I and S+I rNFL bundles were 1.017 ± 0.8427, 0.722 ± 0.8205 and 1.739 ± 1.508 (respectively) for the glaucoma group, and 3.832 ± 0.941, 3.672 ± 0.8777 and 7.504 ± 1.641 (respectively) for the control group. The volumes of these bundles were 127.3 ± 110.2, 92.2 ± 110.1 and 219.6 ± 199.7 (respectively) for the glaucoma group, and 536.9 ± 151.8, 526.4 ± 152.0 and 1063.0 ± 274.3 (respectively) for the control group. Similar results were observed using the low and high threshold values.

Conclusions: rNFL bundle area and volume analysis may be a promising tool for glaucoma assessment. This study demonstrated decreased area and volume in the S, I and S+I rNFL bundles in glaucoma patients compared to controls. Further research is needed to refine which threshold values will yield the best specificity and sensitivity for glaucoma detection, and how this method compares to current glaucoma OCT analyses.