
The widespread functional network of interactions between brain regions is a key aspect to understand brain function. Although it has been investigated with several techniques; neuroscientists are still looking for methods that summarize this complex topology in a simple but informative way.
After estimating functional brain networks from fMRI signals, we use the recently introduced technique of eigenvector centrality to identify the key nodes of this network. We applied this technique to recordings of spontaneous activity in rat, monkey and human. The results identify a robust set of regions involving both primary sensory and associative regions. This might form the elements of a crucial subsystem responsible for passing information to the rest of the network.