Purchase this article with an account.
Brendan McGeehan, Xi Chen, Vincent Tai, Gui-Shuang Ying, Christian Viehland, Ryan Imperio, Katrina Winter, William Raynor, Du Tran-Viet, Shwetha Mangalesh, Maureen Maguire, Cynthia Toth; Repeatability and reproducibility of axial and lateral measurements on handheld optical coherence tomography systems. Invest. Ophthalmol. Vis. Sci. 2020;61(7):1870.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Despite the use of handheld optical coherence tomography (OCT) in many clinical studies to assess the pediatric retina, the repeatability and reproducibility of handheld OCT systems has not been evaluated to date. Here we compare the repeatability and reproducibility of axial and lateral measurements using a handheld spectral domain OCT system (Envisu, Leica microsystems), a research handheld swept source OCT system (UC3), and a tabletop spectral domain OCT system (Spectralis, Heidelberg).
10 healthy adult volunteer subjects were imaged 5 times with each of the three systems by 2 imagers. Two graders measured central foveal thickness (CFT) and 3 graders measured optic nerve to fovea distance (OFD). CFT (µm) is measured between the internal limiting and Bruch’s membranes using customized MATLAB scripts. OFD (mm) is measured between the fovea and the center of the optic disc. A components of variance analysis identified variability due to system, grader, imager, subjects, and random error. Repeatability is the variance due to random error. Reproducibility is the sum of the variance components of random error, imager, and grader. Variation is reported as standard deviation (SD).
The mean (SD) across systems was 227.3 (16.1) for CFT (µm) and 4.375 (0.372) for OFD (mm). The mean (SD) CFT (µm) was 223.0 (14.7), 223.0 (13.9), and 236.0 (15.8) for Envisu, Spectralis, and UC3 respectively. For OFD (mm), the corresponding means (SD) were 4.414 (0.401), 4.383 (0.336), 4.330 (0.378). Across and within systems, the smallest sources of variation were due to imager and grader (see Table 1 and Table 2). For CFT overall (µm), repeatability was 6.2 and reproducibility was 9.7. For OFD overall (mm), repeatability was 0.127 and reproducibility was 0.136. Variation for each component was least for Spectralis except imager variation, which was smallest for UC3. Within systems, repeatability ranged from 4.7 to 6.9 µm for CFT and 0.072 to 0.128 mm for OFD. The reproducibility within systems ranged from 4.7 to 7.3 µm for CFT, and 0.072 to 0.132 mm for OFD.
Little variation was contributed by different imagers and graders. Spectralis had the best repeatability and reproducibility for both CFT and OFD. Envisu had better repeatability and reproducibility then UC3 for CFT measures, but worse for OFD measures.
This is a 2020 ARVO Annual Meeting abstract.
This PDF is available to Subscribers Only