Purchase this article with an account.
Federico Di Matteo, Julia Lamparter, Ryo Asaoka, Richard Russell, Tuan A. Ho, David F. Garway-Heath; Effect of Ocular Magnification on Retinal Nerve Fibre Layer Thickness Measurements in Fourier-Domain Optical Coherence Tomography. Invest. Ophthalmol. Vis. Sci. 2011;52(14):169. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Commercial optical coherence tomography (OCT) devices measure the RNFL at a fixed angular distance from the disc centre. Consequently, the actual distance is greater in myopic eyes. Current models to correct for ocular magnification assume that the retinal nerve fibre layer (RNFL) cross-sectional area (CSA) under different circular scans is constant. The aim of this study is to investigate the relationship between CSA and scan radius as measured by FD-OCT, and to develop a new model to assess the impact of the magnification effect on RNFL thickness measurements.
We measured the RNFL thickness along six circles (diameters: 4, 3.7, 3.4, 3.1, 2.8, 2.5 mm) in 31 eyes of 19 healthy volunteers using the RTVue NHM4 scan protocol. Actual scan diameters were corrected using the Bennett's method. Multilevel regression of RNFL CSA with corrected scan diameters was used to impute the value of RNFL thickness at the actual 3.45 mm.
A significant correlation between 360º CSA and scan radius was found, both in the uncorrected (multilevel regression coefficient: 54, CI 24-84) and magnification-corrected (multilevel regression coefficient: 57, CI: 34-86) analyses. The CSA under each scan arc showed a significant positive correlation with radius in the superior, temporal and inferior quadrants, and a negative correlation in the nasal quadrant. The mean difference between uncorrected (107 µm) and imputed 360º RNFL thickness at the corrected 3.45 mm distance (104.6 µm) was 1.3% (range 8.4 to -9.5%).
Our results are not consistent with the hypothesis that the CSA under different circles is constant. The overall effect upon the estimation of RNFL thickness at the 3.45 mm scan diameter is relatively small in this unselected sample of healthy subjects. The effect at the extremes of refractive error (and axial length) is more likely to be clinically significant.
This PDF is available to Subscribers Only