A comparison of the performance of three visual evoked potential-based methods to estimate visual acuity
Abstract
PURPOSE: To compare visual acuities estimated by three methods of visual evoked potential (VEP) recordings to those obtained by two subjective measures [ETDRS and FrACT (Freiburg acuity test)]. METHODS: Ten healthy subjects, aged between 26 and 67 years (mean 43.5), were examined.
Best-corrected acuity determined by the ETDRS was between 0.03 and -0.3 logMAR (mean -0.06). Sweep VEPs (sweepVEP), pattern appearance VEPs (pappVEP) and steady-state VEPs (ssVEP) were recorded with two electrode placements (10-20 and Laplace) with best optical correction and with artificially degraded vision using five Bangerter occlusion foils, reducing acuity to about 0.1, 0.22, 0.52, 0.7 and 1.0 logMAR (0.8, 0.6, 0.3, 0.2 and 0.1 decimal scale).
Two runs were performed. RESULTS: ETDRS and FrACT acuities showed good agreement, even though ETDRS seemed to underestimate acuity compared with FrACT at higher acuities.
Laplace derivation did not improve any of the VEP-estimated acuities over the 10-20. SweepVEP tended to overestimate lower FrACT acuities, but showed good repeatability.
PappVEP placed FrACT acuities into correct or neighboring categories in 87 % of cases. Average ssVEP acuity showed little difference to those of FrACT but variance was larger.
ROC analysis for typical clinical application showed good performance for all three methods. CONCLUSIONS: The two subjective measurements of acuities are well correlated.
Under the conditions of our experiment, sweepVEP results were less variable and had a better repeatability than ssVEP acuities, whose analysis, in contrast to sweepVEP, can be automated. PappVEP estimates, however, offer a viable alternative, that is, quicker but of lower performance regarding the detection of low acuity thresholds.
All methods had a good performance regarding minimum acuity detection if an average of two runs is used.