A multimodal brain imaging study of repetition suppression in the human visual cortex

Repetition suppression (RS) relates to a reduced neuronal response to a stimulus that is repeated. This phenomenon has been observed in the visual ventral stream and other sensory modalities, suggesting that it is a common feature of neuronal processing. Whilst a number of different models have been suggested to explain the underlying neural mechanisms of RS, they are difficult to test due to variety in paradigm design and the limited resolution of different measuring modalities. This study combined information from different modalities using the same paradigm across the same subjects, in an attempt to create a clearer link between fMRI, magnetoencephalography (MEG) and behaviour data, and thus better understand the underlying mechanism of neuronal RS. We used an oriented Gabor patch stimulus separated by two possible interstimulus intervals of 200 or 600 ms and two possible orientation combinations: the second patch was consistently vertical combined with the first patch which was either horizontal (DIFF) or vertical (SAME). For the 200 -ms condition only, behavioural data showed a statistically significant impairment in subjects' ability to discern the direction of tilt at the SAME condition compared to the DIFF condition; fMRI showed suppression of the BOLD response, and MEG showed suppression of the peak amplitude. A significant correlation between the suppressed BOLD and MEG signals confirm the neuronal origin of the BOLD suppression effect. Measurements from the 3 modalities suggest that neuronal RS in the visual cortex in the current orientation-driven paradigm can be best explained by an overall reduction in the size of the response of all neurons (fatigue model) or a reduction in the number of neurons responding (sharpening model).