The effects of cross-orientation masking on the visual gamma response in humans.

Electrophysiological recordings in primates indicate that visual gamma contains distinct broad- and narrowband components that reflect different neuronal processes. Evidence suggests that cross-orientation masking of luminance-defined gratings should differentially modulate these two components. To test this we measured the effect of cross-orientation masking on the gamma response in 12 human participants using magentoencephalography (MEG). Although both the amplitude and the frequency of gamma were modulated by the presence of a cross-orientation mask, we failed to find evidence for distinguishable components: both broadband gamma at stimulus onset and sustained narrowband gamma were similarly modulated by mask contrast. However, we could not confirm the presence of masking effects due to mask contrast being confounded with the contrast of the stimulus as a whole. We therefore tested a further 12 participants in a second experiment in which the stimuli were: a plaid stimulus, the two component gratings which formed the plaid and the same two gratings but with Michelson contrast matched to the plaid. We found that gamma amplitude was reduced and gamma frequency increased to the plaid stimulus when compared with the contrast-matched gratings or with the sum of the two component gratings, indicating that visual gamma was indeed modulated by cross-orientation masking. Surprisingly, masking did not affect the pattern-onset evoked response, challenging previous hypotheses that cross-orientation suppression - the phenomenon by which the response to an orientated grating is suppressed by a cross-orientation mask - is driven by feedforward inputs to V1.