Imaging data is also smoothed to remove minor differences in anatomical variability across subjects. Anatomical variability is a constant issue faced in neuroimaging. Different warping algorithms and landmark systems have been proposed to optimize coregistration in different regions of the brain. It is clear, though, that humans show variability in both the structural and functional topography of the brain that cannot be overcome by coregistration. Most researchers naturally focus on group analyses to report common areas of activation. However, this approach fails to capture more idiosyncratic activations. Moreover, in regions of high structural variability and in tasks that localize to variable locations, the group analysis may fail to detect relevant activations. For instance, portions of the fusiform gyrus are responsive to faces, but the precise location varies across subjects (Kanwisher, McDermott, & Chun, 1997). In such a situation, group analyses could easily fail to detect the presence of this region.
Was this article helpful?