VER amplitude measurements can be used to predict a subject's acuity. The technique involves presentation of a series of checkerboard stimuli with supra-threshold check sizes and plotting a best fit line for VER amplitudes produced by these stimuli. This line is then extrapolated to zero VER amplitude to determine the acuity. Unfortunately, VER amplitude variability can interfere with this procedure and reduce acuity measurement accuracy.
In this study, VER amplitude variability was incorporated into a computer model of the acuity measurement technique in order to assess its effect on measurement reliability. For each of several different input parameters, the model produced 1,000 acuity measurement simulations and then used the standard deviation of these distributions to evaluate reliability. Several factors influencing the technique were studied, including signal-to-noise ratios, the number of VERs e.vet-aged per check size, and the number of check sizes used to make the prediction.
The results indicate that for a small percentage of cooperative adult subjects (and if certain measurement requirements are met) VER based acuity determinations can be within one Snellen line of actual acuity 95% of the time. For average subjects, acuity predictions will be within one Snellen line of actual acuity 65% of the time, and, for some subjects with low signal-to-noise ratios, the predicted acuity is within one Snellen line only 30% of the time. The important factors which influence acuity measurement reliability are the subject's signal-to-noise ratio and the number of VER amplitudes that are averaged at each check size before plotting the best fit line. The number of check sizes used in the prediction was found to be irrelevant.
Files are restricted to Pacific University. Sign in to view.