Abstract
Purpose :
Eccentric photorefraction and Purkinje image analyses of video are used to estimate refractive state and eye position simultaneously. Beyond their application in screening and studies of typical development, they provide insight into atypical development. To study patients with significant refractive error and spectacles, we must understand their effect on the calibration (e.g. the role of spherical and astigmatic lenses and their vertex distance).
Methods :
A two-dimensional eccentric photorefraction simulation was built in Zemax using sequential ray tracing. The inward component included appropriate light sources to create the retinal image while the outward version created the camera image using the retinal image as the source. The slope of the luminance distribution in the pupil image was plotted against simulated refractive error (-15D to +15D) for pupil sizes from 3mm to 7mm. Calibration estimates were then obtained by simulating introducing -8D to +8D trial lenses at the spectacle plane for a number of conditions: i) vertex distances from 3mm to 23mm, ii) spectacle corrected hyperopic refractive errors of +4D and +7D, and iii) uncorrected and corrected astigmatism of 2D and 4D at 4 different axes.
Results :
The results demonstrated slope saturation and then reversal for significant refractive errors (>+4 & <-6D for 5mm pupil) and significant effects of pupil size. Saturation occurred faster at larger pupil sizes as shown previously. The effect of trial lens image magnification was equivalent to up to 1D for -8D to +8D lenses. Hyperopic and astigmatic spectacle corrections at 13mm vertex distance produced additional magnification effects equivalent to up to 1D for -5D to +5D trial lenses.
Conclusions :
The simulated manipulations could cause up to a 1D difference in estimated defocus when performing calibration within the instrument operating range. Refractive errors beyond the operating range saturation points could be significantly underestimated in both hyperopic and myopic directions
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.