Purchase this article with an account.
W. P. Patton, K. A. Muldrew, M. Stevenson, U. Chakravarthy; Taking the Machine Out of the Equation: Measuring Macular Thickness From Different OCT Capture Systems in a Reading Centre. Invest. Ophthalmol. Vis. Sci. 2010;51(13):1791.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To compare manually-assessed retinal thickness measurements on OCT in a reading centre in healthy and diseased eyes captured by different commercially available OCT instruments.
Three different OCT instruments (Stratus OCT [Carl Zeiss Meditec, Inc. Dublin, CA], RTVue [Optovue Corp., Fremont, CA], Spectralis HRA OCT [Heidelberg Engineering, Inc., Heidelberg, Germany]), were used to capture radial line scans (or equivalent) in healthy or diseased eyes. The scans from each system were exported into JPEG format and imported into custom reading centre software, iP Reading Centre (Digital Healthcare, Cambridge UK). Scans were calibrated in the vertical plane and scrutinised side by side to identify common features. A single scan from each of the different capture systems was chosen at the same orientation and manual measurements were taken from the inner limiting membrane to the visible pathology or the RPE-choroicapillaris complex, whichever was more pronounced.
Twelve eyes were included in the study. Using each patient-eye as a block effect to test each machine, a 2-way ANOVA showed that there were no significant differences between the machine mean measurements of retinal thickness across the three OCT capture systems. The mean thickness measurements for Spectralis, Stratus and RTVue were 0.150, 0.149 and 0.136 respectively.
The three OCT systems provided similar results for retinal thickness measurements. This has important consequences for reading centres and their approach to perform manual measurements for OCT thickness rather than relying on capture system software. This study has several limitations. Although the scan orientation chosen was the same for each system, there can be no guarantee that the same location was used in each image to perform the measurement, or that the scan intersects the same anatomical region in all systems. However, the images were scrutinised in advance to determine common features and the thickness lines were drawn on all scans before the measurements were known. The sample size is also quite small which may result in a type II error and thus the study will be extended to a larger set of patients and will include reproducibility assessments.
This PDF is available to Subscribers Only