Feature Of The Week 9/2/12: Academic Medical Center at The University of Amsterdam Investigates Comparison of RNFL Thickness Measurements by SD-OCT Systems Using a Phantom Eye Model
We developed a multi-layered phantom eye model that distinguishes instrumental from biological variation, to quantify differences in retinal nerve fiber layer (RNFL) thickness measurements by various spectral-domain optical coherence tomography (SD-OCT) systems. The eye model comprised a glass lens, water as vitreous and silicone elastomer based layers mimicking the retina (top layer thickness = 49 µm). We tested 3 RTVue OptoVue, 3 Cirrus Zeiss, 3 3DOCT-1000 Topcon and 3 Spectralis Heidelberg SD-OCT systems. With each system the layered phantom eye model was measured 5 times. An average RNFL thickness and standard deviation were determined by the apparatus using a circular scan (Ø=3.4mm) around a fixed location in the eye model. We used analysis of variance (ANOVA) test to compare means for intra-brand variability and inter-brand variability. Intra-brand variability: RNFL thickness measurements are significantly different for one Topcon and one RTVue and one Cirrus.>br>
Inter-brand variability: RNFL thickness determined with RTVue and Topcon differ significantly with Cirrus and Spectralis. All 12 OCT systems calculated the RNFL thickness on average 13-23 µm thicker than the real thickness of the phantom layer. No OCT system determines the geometrical thickness of the RNFL of the phantom eye. The repeatability of the systems is high.
For more information see recent Article. Courtesy of Roy de Kinkelder from the Academic Medical Center at The University of Amsterdam.