Purchase this article with an account.
Lei Shi, Jim W.L. Lewis, Ying-Ling Chen; Astigmatism measurement in photorefraction. Invest. Ophthalmol. Vis. Sci. 2018;59(9):4763. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Eccentric photorefraction (EPR) is a technique for acquiring objective estimations of refractive errors. With a Badal optometer, EPR could achieve high sensitivity and accuracy for refraction in a selected meridian. The purpose of this study is to experimentally test the accuracy of astigmatism measurements using EPR with a three-meridian/eccentricity scan and a second-order wavefront analysis.
We use a prototype infrared autorefractor that employs EPR and a Badal optometer. The refraction of the eye is measured by moving the EPR camera behind the Badal lens until the pupil’s reflex pattern reverses direction like the retinoscopy. The reversal position of the EPR camera corresponds to the eye’s refraction on the EPR eccentricity meridian. Three scans are performed with lightings along meridians with 120 degree separations. We tested the prototype on our own eyes with contact lenses and on an artificial eye with adjustable refractive error from -7 to +6 diopter and also additional cylinder trial lens from -3 to +3 diopters in 15 degrees intervals. A second-order aberration wavefront model is applied for the recovery of sphero-cylindrical refraction analysis.
We found that the corneal reflection of the artificial eye presents a significant noise and error source which do not exist for human tests. With a simple code that identifies and then discount the reflection pixels, the result with the artificial eye, plus trial cylinder lenses of +3D and -3D showed an overall accuracy (mean error +/- standard-deviation) for (sphere, cylinder, axis) as (0.11D +/-0.18D, 0.15D +/-0.21D, -1+/- 3 degree). The accuracy from 1D cylinder test is further improved in all three results: (-0.10D +/-0.08D, 0.01D +/-0.11D, -0.5 +/- 3 degree).The testing on non-cyclopleged human eyes showed repeatability with variation less than 0.25 diopter and the similar accuracy of error <4 degrees in astigmatic axis.
The EPR optometer with the 3-meridian-scan and the 2nd-order wavefront model analysis are reliable in detecting astigmatism in our study. Image processing to remove the corneal reflection is a necessary procedure to perform the analysis in artificial eye images.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.
This PDF is available to Subscribers Only