Purchase this article with an account.
Samantha Slotnick, Ivan Bodis-Wollner; Veridicality of SD-OCT in Oblique Retinal Scans. Invest. Ophthalmol. Vis. Sci. 2011;52(14):1308.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To determine whether SD-OCT data output as provided by two manufacturers is influenced by the method of measuring thickness.
Myopic retinas were selected for study due to oblique on-screen retina display, 2° to off-axis beam penetration in prolate globes. Numerical & image data from macular scans provided by Cirrus HD-OCT Review Software 5.1 (Zeiss, (Z)) & by ReVue Research Browser 6.1 (Optovue) are analyzed & compared. Z software outputs retinal thickness (ILM-RPE), ILM height (ILMh) & RPE height (RPEh). On-screen image (OSI) analysis is compared with these data. Rotational transformation of ILMh & RPEh is performed in order to set the RPE to horizontal and measure thickness normal to the RPE. Data from RTVue EMM5 scans are manually extracted at line scan intersection points in a 17x17 array & compared with OSI measurements obtained with ReVue.
The Z retinal thickness (ILM-RPE) was determined to measure thickness on a vertical axis through the screen, but not normal to the RPE. The difference between ILMh & RPEh was approximately half (ILM-RPE): 2*(ILMh-RPEh)-(ILM-RPE)= 6.5±1 um. OSI analysis of Z's 1:2 ratio image obscures accurate clinical observation of sections measured normal to the retina (see image). RTVue thickness data are within reported axial resolution error (5um) whether measured normal to the RPE or vertical to screen, but a cumulative error in transverse localization is introduced on oblique scans, increasing with foveal eccentricity.
1. For research applications as well as for comparability of data across different tests it would be helpful if manufacturers stated precisely how the thickness measurements are obtained. If the accurate location of the ILM & RPE are provided, data could be transformed (for comparative research purposes) & thickness measurements could be obtained normal to the RPE & at precise foveal eccentricities. 2. OSI display in 1:1 ought to be available, enabling veridical clinical assessment of thickness measurements normal to the RPE for both clinical & comparative research purposes.
This PDF is available to Subscribers Only