Purchase this article with an account.
Jason D Marsack, Lingyan Anderson, Ayeswarya Ravikumar, Heather A Anderson, Jos J Rozema, Raymond A Applegate; Comparing Visual Image Quality with Objectively and Subjectively Derived Spectacle Corrections in Subjects with Keratoconus. Invest. Ophthalmol. Vis. Sci. 2014;55(13):2714.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Subjective refraction is typically challenging in patients with keratoconus (KC) due, in part, to the presence of elevated higher-order aberrations (HOA). However, aberrations interact, providing an opportunity to use sphere and cylinder to optimize visual image quality in the presence of elevated HOA. The purpose of this study was to objectively search for the residual sphero-cylindrical error that maximized visual image quality in the presence of uncorrected HOA in KC and to compare resultant visual image quality and expected gain in acuity to that achieved with subjectively prescribed spectacles.
Uncorrected 5mm wavefront error was previously obtained for right eyes of 56 KC subjects. Third - 8th order aberration data were combined with 8177 unique residual sphero-cylindrical error conditions within ±2.00D of sphere (0.25D steps) and up to -2D of cylinder (0.25D steps) of a full lower-order (sphero-cylindrical) correction. Resulting Zernike descriptions were used to identify the maximal objectively derived value of the visual image quality metric visual Strehl ratio (VSX). Habitual sphero-cylindrical corrections (available in 10 of the eyes) were also mathematically applied to the 2nd - 8th order uncorrected aberration data, and the resulting data utilized to calculate subjectively derived VSX.
In the 22 eyes where VSX achieved a maximal value within the range of sphero-cylindrical conditions studied, the objectively derived VSX value was better than the value with zeroed lower-order terms (p<0.0015) with median maximized VSX values of 0.189 (1Q: 0.091; 3Q:0.227) and 0.041 (1Q: 0.015; 3Q:0.075) respectively. Likewise, the refraction defined objectively produced larger values of VSX (p < 0.002) in the 10 eyes than the habitual spectacles, where median maximized VSX was 0.058 (1Q: 0.049 3Q: 0.081). The predicted mean gain in VA with the objective vs. subjective correction based on models in the literature was 2.5 lines.
VSX values achieved with this optimization were better than those achieved by simply zeroing lower order terms or those obtained with subjective refraction. These data suggest visual performance for KC patients wearing spectacles might be improved if they are prescribed objectively, using metrics predictive of visual performance, considering a large number of possible residual lower- order error conditions.
This PDF is available to Subscribers Only