Abstract
Purpose: A prominent feature of many recordings of stimulus induced field potentials in visual cortex is a narrowband increase in power centered within the gamma band (30-100 Hz). The narrowband peak has attracted a great deal of attention, with suggestions that it may reflect neural activity critical for fundamental perceptual and cognitive processes. However, recent studies of local field potentials in cat and macaque have shown that the characteristics and even presence of the induced narrowband response depend on the image spatial structure, indicating that this response may not be essential for seeing. We tested whether stimulus-induced narrowband gamma responses can be measured in human visual cortex using electrocorticography (ECoG), and whether the presence of these responses is dependent on the image spatial structure. Methods: A clinical subject with subdural patch electrodes (2.3 mm diameter) was presented with static images for 500 ms each (1000 ms ISI). Seven classes of stimuli were presented (30 exemplars per class), including high contrast vertical gratings (0.16, 0.33, 0.65, or 1.3 cpd square wave) and noise patterns (1/f^4, 1/f^2, and 1/f^0 spectral power distributions). Results: ECoG responses to the images were measured in 11 electrodes adjacent to V1 or on the V1/V2 boundary. The temporal ECoG response could be separated into (1) a broadband spectral increase of up to 200% above baseline, spanning frequencies from <20 Hz to >150 Hz, and (2) a gamma band increase of up to 1500% above baseline, peaked between 30 and 50 Hz. The noise patterns only elicited the broadband response, whereas the grating stimuli elicited both the broadband and the narrowband gamma responses. Conclusions: Narrowband gamma responses can be induced in human visual cortex and reliably measured with ECoG. These responses arise in response to certain types of visible image patterns, but not in response to all images.
Meeting abstract presented at VSS 2013