Abstract
Purpose: :
Visual acuity for letter optotypes is often assumed to be mediated by constant object spatial frequency information (cycles per letter [cpl]), regardless of the observer’s visual acuity. The purpose of this study was to test this conventional assumption by measuring visual acuity using the tumbling E optotype in a group of subjects with a wide range of acuity values.
Methods: :
Subjects included 10 visually normal individuals, ages 46 to 64 years; 4 patients with diabetic retinopathy, ages 58 to 71 years; and 4 patients with age-related macular degeneration, ages 65 to 74 years. Threshold log MAR (minimum angle of resolution) was measured using a set of briefly presented tumbling Es that were either unblurred or blurred with Gaussian low-pass filters that had cut-off frequencies spanning a range of 1.2 log units. The data were fit with an equivalent blur model to estimate the object spatial frequencies that govern visual acuity for the unblurred optotype.
Results: :
Threshold log MAR values for these subjects ranged from -0.1 to 0.6 (20/16 to 20/80 Snellen equivalent). There was a statistically significant correlation between threshold log MAR and log cpl (r = 0.87, p < 0.001), such that subjects with higher values of threshold log MAR used higher object spatial frequencies to judge the orientation of the tumbling E. By reducing the contrast of the tumbling E for normal subjects, both the threshold log MAR and cpl values were increased, thereby simulating the findings in patients. This simulation provides evidence that the relationship between object spatial frequency and visual acuity does not depend on the source of visual acuity loss.
Conclusions: :
Contrary to the standard assumption, the object spatial frequencies mediating visual acuity are dependent on the angular subtense of the optotype at the acuity limit, which poses a challenge in relating visual acuity for broadband optotypes to retinal spatial frequency in cycles per degree (cpd).
Keywords: visual acuity • perception