The relationship between the optical density of cataract and its influence on retinal nerve fibre layer thickness measured with spectral domain optical coherence tomography

Abstract

The purpose of this study was to model the influence of cataract on Spectral Domain Optical Coherence Tomography (SDOCT) image quality and Retinal Nerve Fibre Layer (RNFL) thickness measurements. SDOCT images, made with two different devices (3DOCT-1000, Topcon and Cirrus HD-OCT), before and after cataract surgery were compared and judged against measurements from normal subjects using artificial filters simulating the effects of cataract. Optical density of the images was calculated based on a mathematical model described previously. In total, forty-eight eyes were included for pre- and postoperative cataract extraction measurements. OCT image quality significantly (p < 0.001) improved postoperative and postoperative RNFL thickness was significantly (p < 0.001) thicker in both groups of patients. The measurements using artificial filters showed a rather precise linear relation between change in filter induced optical density and change in RNFL thickness (R = 0.941, p < 0.001 for 3DOCT-1000 and R = 0.785, p < 0.001 for Cirrus HD-OCT). For the patient groups, the relation was less marked, 3DOCT-1000 Rs = 0.697, p < 0.03 and Cirrus HD-OCT Rs = 0.444, p < 0.03. The predictive potential based on the found linear relationship between OCT-effective optical density of cataract and the cataract-induced underestimation was however limited, and mean difference ± SD between predicted and measured RNFL thickness were 1.68 ± 7.55 (3DOCT-1000) and 3.71 ± 2.97 (Cirrus HD-OCT) micron. A linear relationship exists between OCT-effective optical density of cataract and underestimation of RNFL thickness measured with OCT. This finding holds promise to correct for cataract-induced changes in RNFL measurements, but will differ for each type of OCT device.