The properties of induced gamma oscillations in human visual cortex show individual variability in their dependence on stimulus size

The role of gamma-band (typically 30-100 Hz) oscillations in visual processing is a topic of increasing interest. One hypothesis is that gamma oscillations reflect the action of GABAergic inhibitory processes in the visual cortex responsible for surround-suppression. Evidence from primate neurophysiology [Gieselmann & Thiele, A., 2008. European Journal of Neuroscience 28, 447-459.] suggests that the amplitude of the gamma-band response increases as a visual grating stimulus expands outside of the classical receptive field into the inhibitory surround; with the amplitude of the response increasing, and the frequency of the response decreasing, monotonically with stimulus size. In this study, we tested the relationship between the gamma-band response and the size of visual grating stimuli in humans using MEG. In two initial experiments we found that, while the absolute magnitude of the gamma-band response varied considerably across participants, in all cases the amplitude of the response had a monotonically increasing relationship with size. In contrast, we did not find any relationship between the frequency of the response and the size of the stimulus. Previously, the frequency of the visual gamma-band response has been found to correlate across individuals with the surface area of cortical area V1 [Schwarzkopf et al., 2012. Journal of Neuroscience 32, 1507-12.] We, however, were unable to find any correlation between the frequency or the magnitude of the gamma-band response and the dimensions of V1 cortical gray matter as measured from participants' MR images. Consistent with a saturation of the gamma-band response found for some individuals in the first two experiments, in a third experiment we found that the magnitude of the response to our largest stimulus (8°) was less than that predicted from the response to the stimulus' parts.

► The magnitude of the visual gamma-band response increased with stimulus size. ► Individuals with the strongest response showed the least proportional increase with size. ► In contrast, the frequency of the response was not affected by size. ► Parameters of the response did not correlate with V1 structural measures. ► The responses to different grating patches combined sub-additively.