Cirrus OCT versus Spectralis OCT: differences in segmentation in fibrovascular pigment epithelial detachment

Abstract

Automatically measurements of retinal thickness by optical coherence tomography (OCT) facilitate the assessment of various retinal diseases.The aim of this retrospective study was to report macular thickness measurements in eyes with vascular pigment epithelial detachment (PED) due to age-related macular degeneration (AMD) by using two different commercially available spectral domain (SD) OCT instruments and to consequently point out differences in their algorithm software.systems. OCT images of patients with vascular PED due to AMD, obtained with Cirrus and Spectralis OCT, were retrospectively analyzed. Main objectives were to observe differences in central retinal thickness (CRT) values and failures in automated threshold delineation, as well as central point thickness values obtained after manual correction of threshold lines. Scanning with the Cirrus HD OCT was performed with the 512 × 128 scan pattern; scans performed with the Spectralis OCT were 20 × 15 degree raster scans consisting of 19 high-speed line scans. OCT images of 34 eyes of 28 patients with a mean age of 71 years and a mean distance visual acuity (VA) of 0.70 ETDRS were analyzed. Mean central retinal thickness (CRT) was 262.38 μm ± 133.18 (176-507 μm) in Cirrus and 337.82 μm ± 137.75 (277-790 μm) in Spectralis scans,mainly caused by different software approaches in positioning the posterior threshold line, following the PED in Cirrus OCT whereas remaining unelevated in Spectralis OCT. There were failures in positioning the outer retinal boundary line in 50% of Cirrus scans and in 73.52% of Spectralis scans. We obtained the mean value of central point neurosensory retinal thickness of each central single scan after manual delineation, and found a significant correlation (r = 0.819, p < 0.001). Our study indicates that there are significant differences in CRT values in patients with vascular PED, due to different segmentation algorithms and a high error rate in automatically set threshold lines. When planning and conducting multicenter studies, one has to be especially aware of the differences in delineating threshold algorithm lines by different SD OCT devices.