Metrics of Retinal Image Quality Predict Visual Performance in Eyes With 20/17 or Better Visual Acuity

Abstract

The purpose of this study is to determine the ability of single-value metrics of retinal image quality of the eye to predict visual performance as measured by high (HC) and low (LC) -contrast acuity at photopic (P) and mesopic (M) light levels in eyes with 20/17 and better visual acuity. Forty-nine normal subjects in good health ranging in age from 21.8 to 62.6 with 20/17 or better monocular high-contrast logarithm of the minimum angle of resolution (logMAR) acuity served as subjects. Wavefront error through the 10th Zernike radial order over a 7-mm pupil was measured on each test eye using a custom-built Shack/Hartmann wavefront sensor. For each eye, 31 different single-value retinal image quality metrics were calculated. Visual acuity was measured using HC (95%) and LC (11%) logMAR at photopic (270 cd/m) and mesopic (0.75 cd/m) light levels. To determine the ability of each metric of retinal image quality to predict each type of logMAR acuity (P HC, P LC, M HC, and M LC), each acuity measure was regressed against each optical quality metric. The ability of the metrics of retinal image quality to predict logMAR acuity improved as luminance and/or contrast is lowered. The best retinal image quality metric (logPFSc) accounted for 2.6%, 15.1%, 27.6%, and 40.0% of the variance in P HC, P LC, M HC, and M LC logMAR acuity, respectively. In eyes with 20/17 and better P HC acuity, P HC logMAR acuity is insensitive to variations in retinal image quality compared with M LC logMAR acuity. Retinal image quality becomes increasingly predictive of logMAR acuity as contrast and/or luminance is decreased. Everyday life requires individuals to function over a large range of contrast and luminance levels. Clinically, the impact of retinal image quality as a function of luminance and contrast is readily measurable in a time-efficient manner with M LC logMAR acuity charts.