Abstract
Purpose :
The calculation of the retinal magnification factor (RMF) is a first step towards determining retinal feature sizes in units of length. This is particularly important for the creation of normative databases of adaptive optics (AO) ophthalmoscopy biomarkers, such as cell size, cell density, and vessel diameter. Accurate RMF calculation relies on the individualization of schematic eyes through the incorporation of ocular biometry data. Here we compare four paraxial schematic eyes commonly used in AO retinal imaging, against a new individualized schematic eye, in which each of its four surfaces is estimated from ocular biometry and the media between them is assumed to have homogeneous refractive indices.
Methods :
The anterior and posterior surfaces of the cornea and crystalline lens were imaged with an ocular biometer (HP-OCT, Cylite, Notting Hill, Australia), and fitted to a linear combination of ANSI standard Zernike polynomials up to 10th order. Using the optical path length between each surface for each ray returned by the biometer, index-corrected surfaces were generated using real ray tracing. Real ray tracing through these surfaces was then used to calculate the RMF at the preferred retinal locus of fixation (PRL) in individualized four-surface schematic eyes.
Results :
Across a sample of 23 eyes of 17 subjects and with spectacle prescription in the -5.5D to +1D, the comparison of the paraxial RMF calculations and the proposed RMF differ by us much as 5.0% (mean 1.5%). Importantly, the incorporation of decentering and tilt of the optical surfaces results in RMF changes with orientation that in our dataset are as large as 2.5%.
Conclusions :
A novel RMF calculation method that uses four ocular surfaces estimated using optical biometry is demonstrated and tested in a small subject cohort. The method could be refined by incorporating the refractive index variations within the crystalline lens, whenever these can be experimentally measured in the intact living eye.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.