Optical density filters modeling media opacities cause decreased SD-OCT retinal layer thickness measurements with inter- and intra-individual variation.

Abstract

To assess the effect of media opacities on thickness measurements of the peripapillary retinal nerve fibre layer (pRNFL) and macular inner retinal layer (mIRL) performed with spectral-domain optical coherence tomography (SD-OCT) using a set of filters with known optical density. Spectral-domain optical coherence tomography volume scans of the optic disc and the macular area were performed in 18 healthy volunteers, using Topcon-3DOCT-1000 Mark II. A set of five filters with optical density ranging from 0.04 to 0.69 was used. The correlation was calculated between the percentage change in thickness measurements (%ΔpRNFL and %ΔmIRL) and the change in optical density. All scans and measurements were performed in duplicate by one operator. Eighteen right eyes of 18 healthy volunteers were included in this study. Percentage decrease in pRNFL and mIRL thickness correlated with change in optical density (Spearman's rho r = 0.82; p < 0.001 and r = 0.89; p < 0.001, respectively). The measured decrease in pRNFL thickness differed from the decrease in mIRL thickness, not only between individuals, but also within the same individual. Optical coherence tomography thickness measurements of both pRNFL and mIRL are influenced by image degradation caused by optical density filters as a model for media opacities. An underestimation of the thickness of these layers was observed, caused by a shift of retinal layer boundary placement due to image quality loss. This underestimation is not the same for each individual and also differed between the pRNFL and mIRL thickness measurements. These individual and interindividual differences demonstrate that an individual approach will be necessary to correct for this underestimation per layer.