Table 3.
Comparison of average VEP SF limits and behavioural acuity (recognition based on a discrimination task) in normally sighted adults
| Study | Number of subjects | VEP stimulus | Average VEP SF limit | Average acuity (detection task) | VEP minus behavioural difference (log units) 95% LoA |
VEP method | Discrimination acuity method |
|---|---|---|---|---|---|---|---|
| #321 [12] | 40 | On/offset (40/93 ms, 7.52 Hz) checkerboard, 40%, 45 cd m−2 |
26.7 cpd 0.051 logMAR |
− 0.241 logMAR |
0.290 − 0.010 to 0.590 |
Monocular, Laplacian around Oz, DFT, automated extrapolation of significant, noise-corrected magnitudes versus log SF to 0 μV | Freiburg acuity test (FrACT) with Landolt Cs, monocular |
| #47 [78] | 16 | Reversing (6 rps) checkerboard, 88%, 72 cd m−2, 8.4 × 6.5°–16.8 × 13° |
16.2 cpd 0.297 logMAR |
− 0.211 logMAR |
0.508 0.278 to 0.693 |
Monocular, Oz–Fpz, time domain and frequency domain, extrapolated curvilinear or linear amplitude versus checksize | Landolt C bracketed “walk-back”, forced-choice technique, 8/10 correct, 246 cd m−2, 79%, monocular, repeated × 3 |
| #54 [88] | 16 | Vertical sinusoidal gratings, 80%, 80 cd m−2, 15 rps |
37.5 cpd − 0.097 logMAR |
− 0.077 logMAR | − 0.020 | Monocular, Oz–O1/O2, DFT, linear extrapolation of magnitude to 0 μV | Projected tumbling Es, descending method of limits, 4 alternative forced-choice, monocular |
| − 0.144 logMAR | 0.047 | Flom S-chart, Landolt Cs and tumbling Es, 4 alternative forced-choice, 50% correct from psychometric function, monocular | |||||
| #262 [70] | 13 | on/offset (40/400 ms, 2.3 Hz) checkerboard, “350 lx”, 2 × 2° |
24.8 cpd 0.083 logMAR |
≤ 0.000 logMAR | “Typically 0.155″ | Monocular, Laplacian around Oz, 50 averages, subjectively judged reproducibility, amplitude versus checksize plot, linear regression through “clear descending trend” to 0 μV | Landolt C chart at 5 m, 350 lx, monocular |
| #324 [19] | 13 | Vertical sinusoidal gratings, reversing (20 rps), 30%, 132 cd m−2, 20 × 20° |
21.8 cpd 0.135 logMAR |
− 0.162 logMAR |
0.289 0.127 to 0.559 |
Monocular, optimal electrode from 64-channel array, time domain averaging–DFT, minimal SF with significant, noise-corrected magnitude | FrACT logMAR tumbling Es, monocular |
| #252 [101] | 9 | Reversing (2 rps) checkerboard, 20 or 40%, 42 or 10 cd m−2, 4.5 × 4.5° |
~7 cpd ~0.63 logMAR |
0.000 logMAR | 0.63 | Monocular, Oz–earlobe, time domain averaging, minimal SF with subjectively judged VEP | Landolt test chart, monocular (data extrapolated only for subjects with decimal acuity 1.0) |
| #322 [16] | 2 | On/offset (40/93 ms, 7.52 Hz) checkerboard, 40%, 45 cd m−2 |
26 cpd 0.062 logMAR |
− 0.182 logMAR | 0.247 | Monocular, Laplacian around Oz, DFT, automated extrapolation of significant, noise-corrected magnitudes versus log SF to 0 μV | Uncrowded FrACT, Landolt Cs, monocular, repeated |
| #39 [28] | 1 | Reversing (12 rps) checkerboard, 60%, 80 cd m−2 |
19.37 cpd 0.190 logMAR |
0.027 logMAR | 0.160 | Monocular Oz–Fz, DFT, linear extrapolation of magnitude versus SF to 0 μV | Landolt C chart, monocular |
| #222 [44] | 1 | Vertical sinusoidal gratings, on/off (sinusoidal, 16 Hz), 40%, 17 cd m−2, 5 × 5° |
20 cpd 0.176 logMAR |
− 0.06 logMAR | 0.237 | Oz–Cz, time domain averaging–DFT–vector averaging, extrapolation | Landolt acuity |
LoA: 95% limits of agreement (see Fig. 5b). Hashtag numbers indicate the 155 references whose data were included in the systematic review