Axial Length–Based Ocular Magnification Correction Improves the Association Between Optic Disc and Retinal Nerve Fibre Layer Measurements

Abstract

Purpose: : Eye–camera magnification corrections that use axial length measurements are more accurate than those that don’t (Garway–Heath BJO 1998). We investigated the influence of eye–camera magnification correction on the relation between optic disc area (DA), rim area (RA) and retinal nerve fibre layer thickness (RNFLT) measured by HRT–II and Stratus–OCT.

Methods: : DA and RA measured by HRT–II were retrospectively obtained from 74 normal subjects, 37 males and 37 females, mean age 64.1±10 years. From 38 of these subjects (17 females, 21 males, mean age 63.3±11.6) RNFLT measurements had been obtained by Stratus–OCT (RNFLT 3.4 scan, RNFLT single–eye analysis protocol). DA and RA were calculated in the HRT standard software (1.4.1.0) and in a version modified to correct eye/camera magnification by axial length measurement. RNFLT was calculated in the OCT software and recalculated in a spreadsheet with an axial–length correction algorithm (assuming a linear decline of RNFLT with distance from the optic disc centre). The relation between age, DA, RA and RNFLT was explored by linear regression analysis.

Results: : : Mean axial length was 23.62±1.34 mm (mean±SD). Mean DA was 1.79±0.38 mm² and 1.81±0.44 mm² after magnification correction. Mean RA was 1.35±0.22 mm² and 1.37±0.26 mm² after correction. RNFLT was 91.6±9.4 µm and 88.2±9.3 µm before and after correction, respectively.The relation between DA and RA was moderate and improved after axial length magnification correction (r²=0.28, p<0.0001 before, and r²=0.4 after correction p<0.0001). The relation between DA and RNFLT was not significant either before or after correction (r²=0.02, p=0.4 before and r²=0.07 after correction p=0.1). The relation between RA and RNFLT was fair (r²=0.13, p=0.039) and improved after axial length magnification correction (r²=0.26, p=0.002). The relation between RNFLT and age was fair (r²=0.16, p=0.012) and improved after correction (r²=0.27, p=0.001), while RA and age were poorly related, either before and after correction (r²=0.007, p=0.4 and r²=0.027, p=0.1).

Conclusions: : More accurate correction of the eye/camera magnification error improves the relationship between RA and both DA and RNFLT and between RNFLT and age. RA was not related to age and RNFLT was not related to disc area, either before or after magnification correction, in this small cohort.

Keywords: optic disc • imaging/image analysis: clinical