Visual evoked potential-based acuity assessment: overestimation in amblyopia

Abstract

When visual acuity (VA) is assessed with spatially repetitive stimuli (e.g., gratings) in amblyopes, VA can be markedly overestimated. We evaluated to what extent this also applies to VEP-based objective acuity assessment, which typically uses gratings or checkerboards. Seventeen subjects with amblyopia (anisometropic and strabismic) participated in the study; decimal VA range of their amblyopic eye covered 0.03-1.0 (1.5-0.0 logMAR). Using the Freiburg Acuity VEP (FrAVEP) method, checkerboard stimuli with six check sizes covering 0.02°-0.4° were presented in brief-onset mode (40 ms on, 93 ms off) at 7.5 Hz. All VEPs were recorded with a Laplacian montage. Fourier analysis yielded the amplitude and significance at the stimulus frequency. Psychophysical VA was assessed with the Landolt-C-based automated Freiburg Visual Acuity Test (FrACT). Test-retest limits of agreement for both FrACT and FrAVEP were ±0.20 logMAR. In all but two dominant eyes and high-acuity amblyopic eyes (VA<0.3 logMAR), FrACT and FrAVEP agreed within the expected limits of ±0.3 logMAR. However, the VEP-based acuity procedure overestimated single Landolt-C acuity by more than 0.3 logMAR in 9 of 17 (53%) of the amblyopic eyes, up to 1 logMAR. While all subjects had a psychophysical acuity difference>0.2 logMAR between the dominant and amblyopic eye, only three of them showed such difference with the FrAVEP. Both measurements of visual acuity with the VEP and FrACT were highly reproducible. However, as expected, in amblyopia, acuity can be markedly overestimated using the VEP. We attribute this to the use of repetitive stimulus patterns (checkerboards), which also lead to overestimation in psychophysical measures. The VEP-based objective assessment never underestimated visual acuity, but needs to be interpreted with appropriate caution in amblyopia.