Abstract
Purpose :
To determine whether using the image quality (IQ) metric VSX (visual Strehl ratio) to optimize objective refraction from wavefront error (WFE) measures can provide a refraction as good as, or better than, subjective refraction while undilated or dilated.
Methods :
14 subjects, mean (±SD) age 27.6 (±5.3) years participated; all had refractive error >1D (range +1.25 to –7.00D sphere and 0 to –1.25D cylinder), correctable to 20/20, and no ocular pathology.
Uncorrected WFE was measured undilated; autorefraction and subjective refraction were performed. For each eye >15 000 sphero-cylindrical combinations (S-C Rx) (from +3 to –2D sphere in 0.25D steps, 0 to –2D cylinder in 0.25D steps, and 2° axis steps, around the second order defocus term) were converted to Zernike terms and added to the measured WFE, each generating a residual WFE. VSX was calculated for each resultant WFE and the S-C Rx with best VSX was identified as optimizing visual IQ. High (HC) and low contrast (LC) visual acuity (VA) were recorded (3 times and averaged) through each Rx. Subjects also evaluated their distance vision preference between the two Rxs on a seven-point Likert scale.
Pupils were dilated with 1% tropicamide and the entire procedure repeated.
Results :
Both eyes of each subject were analyzed. Mean (±SD) difference (objective – subjective) in logMAR VA between the Rxs was –0.003 (±0.037) and –0.026 (±0.081) for HC and LC respectively for the undilated condition, and +0.003 (±0.070) and –0.003 (±0.088) for the dilated condition. Given test-retest variability (SD) of both HC and LC logMAR VA is > ±0.05, the Rxs provide clinically equivalent VA. Paired t-tests also showed no statistically significant difference between the Rxs.
Mean (±SD) differences (objective – subjective) in terms of M, J0, and J45 respectively were –0.62 D (±0.39), +0.08 D (±0.17), –0.03 D (±0.13) undilated, and –0.32 D (±0.23), +0.02 D (±0.16), –0.05 D (±0.12) dilated. The Euclidean separation of the Rxs in power vector dioptric space was 0.68 (±0.35) undilated and 0.40 (±0.21) dilated.
Preferences (objective better: no difference: subjective better) were 19:0:9 for undilated eyes and 15:6:7 for dilated eyes.
Conclusions :
Optimizing objective refraction using VSX provided an Rx equivalent to subjective refraction in VA and generally preferred in a monocular comparison. The two Rxs were virtually identical in astigmatic components and similar in equivalent sphere.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.