Abstract
Visual acuity (VA) is an important clinical measure that defines the visual system’s ability to discern fine details, and is typically administered by a clinician using fixed-range chart-tests, e.g. Early-Treatment-of-Diabetic-Retinopathy-Study (ETDRS). Recently, the response-adaptive, generalizable, self-administered AIM (Angular Indication Measurement) method was introduced (Skerswetat, Boruta, Bex, 2022). Here, we compared AIM-Acuity generated VA, testing duration, and sensitivity to defocus and astigmatic blur with ETDRS results in 23 adults and investigated AIM-Acuity’s test-retest reliability. Defocus and astigmatic blur levels were induced using plano/positive spherical and cylindrical lenses (0.25D-accommodation, ±0.00D, +0.25D, +0.50D, +0.75D, +1.75D, and +0.50D, +1.00D, +2.00D each at 0°, 90°, 135°, respectively) in the dominant eye at 4m viewing distance. Each of the two computer-generated AIM-Acuity charts comprised a 4×4 grid of 6°/cell, surrounded by an indication-ring, containing a Landolt-C with randomized gap-orientations and log-scaled sizes, adaptive across charts. Participants clicked on the indication-ring that corresponded to the perceived gap orientation. Angular orientation errors (indicated-vs-actual difference) were fit with cumulative Gaussian functions to estimate VA. A plano lens was used for the retest condition. A 2-way-ANOVA showed a main effect for defocus-blur (p<0.001) but not for test type. No interaction was found. A 3-way-ANOVA for the astigmatism experiment revealed main effects for test type (median VA across blur-levels: AIM (0.22 logMAR), ETDRS (0.10 logMAR); p<0.01), blur (VA decreased with increasing astigmatism; p<0.001), and direction (oblique-loss>other directions; p<0.001) and no interactions. AIM-Acuity and ETDRS had comparable test durations (median=58.5s, 63.0s, respectively; p>0.05). Bland-Altman-analysis showed small bias, no systematic learning effect, and improvement with more adaptive steps for AIM-Acuity. In conclusion, the self-administered AIM-Acuity showed good test-retest reliability, was comparable in test duration to ETDRS, detected VA-loss due to defocus comparably to ETDRS, and deteriorated more using astigmatic blur.