Dioptric blur is not fully reflected by VEP-based visual acuity estimates

Objective estimation of visual acuity (VA) based on visual evoked potentials (VEPs) has become an established technique for cases where psychophysical VA might be unreliable. Refractive errors and improper accommodation could undesirably affect the outcome of VA measurements. Consequently, it is of interest whether a VA reduction due to dioptric blur is reflected by VEP-based estimation of VA.

We degraded vision in 19 participants to nearly 1.0 logMAR by using either plus lenses or a filter that creates Gaussian blur. For both types of degradation, we compared the outcomes of objective VEP-based VA testing to standard psychophysical VA. For comparison, we also obtained psychophysical grating VA.

With Gaussian blur, both values, VEP-based VA and psychophysical Landolt-C VA, were nearly identical. With dioptric blur, VEP-based VA was better than psychophysical Landolt-C VA in all participants by an average of 0.37 logMAR with some interindividual variability. Psychophysical grating VA was only relatively mildly affected by blur with no sizable differential effect of blur type.

VEP-based estimation of VA does not reveal the full amount of VA reduction in the case of dioptric blur. On the one hand, this decreases VEP-based methods' susceptibility to incorrect refraction and mis-accommodation, which are not normally the targeted causes of VA reduction. On the other hand, it reduces the accuracy in quantifying refraction-related impairments of vision with VEPs.