VEP-based acuity estimation: unaffected by translucency of contralateral occlusion

Abstract

Purpose Visual evoked potential (VEP) recordings for objective visual acuity estimates are typically obtained monocularly with the contralateral eye occluded. Psychophysical studies suggest that the translucency of the occluder has only a minimal effect on the outcome of an acuity test. However, there is literature evidence for the VEP being susceptible to the type of occlusion. The present study assessed whether this has an impact on VEP-based estimates of visual acuity. Methods We obtained VEP-based acuity estimates with opaque, non-translucent occlusion of the contralateral eye, and with translucent occlusion that lets most of the light pass while abolishing the perception of any stimulus structure. The tested eye was measured with normal and artificially degraded vision, resulting in a total of 4 experimental conditions. Two different algorithms, a stepwise heuristic and a machine learning approach, were used to derive acuity from the VEP tuning curve. Results With normal vision, translucent occlusion resulted in slight, yet statistically significant better acuity estimates when analyzed with the heuristic algorithm (p = 0.014). The effect was small (mean ΔlogMAR = 0.06), not present in some participants, and without practical relevance. It was absent with the machine learning approach. With degraded vision, the difference was tiny and not statistically significant. Conclusion The type of occlusion for the contralateral eye does not substantially affect the outcome of VEP-based acuity estimation.