Abstract
Although the amplitudes of steady-state visual evoked potentials (SSVEP) are frequently used to quantify neural responses, past work has often neglected their phases. However, neural latencies generally decrease as response amplitudes increase (e.g. Albrecht, 1995), suggesting that SSVEP phases could contain valuable information. We measured SSVEP amplitudes and phases in response to gratings that varied in contrast, both when presented alone and when presented superimposed on a second, constant-contrast "masking" grating. We recorded SSVEP signals from 4 electrodes placed over the occipital lobe while participants (n=8) viewed 1.5 sec stimulus presentations. A vertical grating phase-reversed at 15 Hz and varied between 3.125-25% contrast in logarithmic steps. In the no-mask condition, the vertical grating was presented alone, while in the masked condition, it was superimposed on a horizontal grating that had a fixed 25% contrast and phase-reversed at 6 Hz. The average 15 Hz SSVEP amplitude in the no-mask condition increased with contrast for the lowest contrast levels, but remained flat at high contrast levels. The phase of the 15 Hz response, however, advanced monotonically with increasing contrast (p < 0.01), indicating a steady decrease in neural latency. In the masked condition, increasing contrast elicited monotonic increases in the 15 Hz SSVEP amplitude, as well as monotonic advances in phase (both p < 0.01). Comparing masked and no-mask conditions showed that adding the second grating both reduced the amplitudes of the 15 Hz response at low contrasts (p < 0.01) and delayed their phases (both p < 0.01). These results are consistent with the masker weakening and slowing neural response. SSVEP phases, at least in the no-mask condition, appeared more sensitive to changes in effective stimulus strength than SSVEP amplitude. This sensitivity should allow future analyses to test specific models of the neural computations underlying visual masking.
Meeting abstract presented at VSS 2017