Refractive error determination based on meridional targets and aberrometry

Abstract

Purpose : To evaluate an objective method of refractive error determination based on aberration measurement and meridional targets.

Methods : Targets consisted of three or more Gaussian-attenuated luminance gratings (sinusoidal or square wave) at uniformly spaced orientations. Simulated eyes were defined by sets of Zernike coefficients (2^nd through 6^th order). Two sets of 50 simulated eyes were generated, one set each for low- and high-aberration eyes. Virtual refractions were done by finding the spherical dioptric power to maximize the neural CSF-weighted contrast of each of the gratings in the Fourier domain. The spherical powers at three or more grating orientations identify the required refractive correction in the meridians orthogonal to the gratings. A one-cycle sinusoid, fit to those spherical powers as a function of meridian, represents the sphero-cylindrical (Sph-Cyl) refractive correction. A second set of virtual refractions were done on the same two sets of eyes using a method that maximizes the magnitude of an image quality metric, the Visual Strehl Ratio (VSX).

Results : The method produces Sph-Cyl corrections very similar to those resulting from the VSX method. In the low aberration set of eyes, the median difference between the two methods was less than 0.075D for both astigmatic (AD) and total dioptric difference (TDD). The 95% percentiles were 0.19D (AD) and 0.26D (TDD). In the high-aberration set of eyes, the median differences were 0.21D (AD) and 0.29D (TDD). The 95^th percentiles were 0.88D (AD) and 1.15D (TDD). Calculated retinal images produced by the Sph-Cyl corrections based on the two methods typically show only small differences in apparent image quality, even for eyes producing the greatest differences in Sph-Cyl correction between the two methods. Computation time needed to find Sph-Cyl corrections by the meridional method is modest compared to that required of the VSX method, generally exceeding a ratio of 50:1 in speed.

Conclusions : This meridional/aberration refraction method can be used as an objective refraction technique. The sphero-cylindrical results of the method are similar to those produced by one that maximizes VSX. The computations required of this method are more concise, primarily because dioptric powers are searched in just one dimension (sphere) rather than in three (sphere, cylinder, axis, or M, J0, J45).

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.