Wolf ME, Maurer JP, Bach M (2005) Objective Estimation of Visual Acuity — High Correlation Across Normal and Diseased Vision with an Automated VEP-Based Algorithm. Proceedings of the 8th Tübinger Perception Conference, p. 110
Purpose: To assess the visual acuity with fully objective evaluation based on visual evoked potentials.
Methods: 40 normal subjects, and 24 patients (10 cataract, 3 corneal clouding, 7 epiretinal gliosis, 3 AMD, 1 chorioretinitis; decimal visual acuity (VA) range 0.14-1.1) participated in the study. Checkerboard stimuli with 6 check sizes covering 0.05-0.4° (or 0.09-0.8° for visual acuity below 0.35) were presented in brief-on-off mode (40 ms on, 93 ms off) at 7.5 Hz. In normal subjects, the stimuli were optically degraded by various degrees of dioptrical defocus (n=35) or Bangerter occluders (n=45) resulting in a decimal VA range of 0.128-2.6. Steady-state VEPs were recorded with a Laplacian montage ((2Oz-(RO+LO)). Fourier analysis yielded the magnitude at the stimulus frequency As and—as noise estimate—the average of the two neighboring frequencies (N). As and N determine the significance level p of the response, and from their ratio the non-noise-contaminated response A* can be calculated. Tuning curves were obtained by plotting A* vs. the dominant spatial frequency of the corresponding checkerboard. A relatively simple fully automated algorithm used the significance level (p<5%) and A* to automatically select an appropriate region in the high spatial-frequency range on which a linear regression was performed. Comparing of several classes of algorithms, the best result was obtained by a stepwise heuristic. "VEP acuity" obtained as the inverse of the zero-extrapolated spatial frequency. Subjective VA was obtained with the automated "Freiburg Acuity Test".
Results: The brief-on presentation evoked high VEP amplitudes. However, many tuning curves displayed the well-known "notch". The analysis algorithm successfully ignored the notch, if present. The resulting VEP acuities, when compared to subjective VA, correlated by r=0.90 and coincided within ±1 octave in 95% in all normals, including the visually degraded runs. In patients, the VEP acuity and subjective VA coincided within one octave in 87.5% of the cases.
Conclusion: The fully automated analysis avoided subjective problems in peak-trough assessment. The slightly lower agreement of VEP acuity and subjective VA in patients indicates that optical degradation does not model all pathologies well, but it was also more difficult for aged patients to operate the automatic subjective acuity test. The results provide quantitative limits to assess patients where subjective testing is impossible or problematic, e.g. in possible malingering.