Brain modeling is certainly a recent field, as fMRI and EEG methods have not become fully mainstream until about 15 years ago, from what I can tell. But most of that brain modeling has been concerned with modeling the brain when it is performing some goal oriented task, such as playing chess or recognizing an object.
Recently, interest in modeling the brain at rest has skyrocketed. The applications are enormous–if we fully understand the brain at rest then we can draw conclusions based upon how it deviates from that state. Ghosh et al have conducted a study regarding that end and attempting to model the brain.
They took a macaque brain activation data from the CoCoMac database, which came in the form of a connectivity matrix with 38 nodes and weights on each ranging from 0 to 3. They then evaluated this data set both from the spatial and temporal perspective, as well as the combination of the two. They found that both of these parameters correlate with the data. As the temporal aspects of the coupling change, the anatomical weights vary.
The model that they develop is interesting, and they claim that it is the first of its kind, since previous models either cannot fit faster wavelength oscillations or make predictions that are too vague. Previous models had also not included temporal factors in their model, which they claim to be essential. Although models without temporal factors can account for BOLD data (where the fluctuations are on the order of less than 0.1 Hz), they cannot account for the faster EEG data.
The authors then go on to use their model in a surprising number of applications, including development hypotheses, making it testable. Certainly an interesting paper on a number of levels.
Ghosh A, Rho Y, McIntosh AR, Kötter R, Jirsa VK 2008. Noise during Rest Enables the Exploration of the Brain’s Dynamic Repertoire. PLoS Computational Biology 4(10): e1000196 doi:10.1371/journal.pcbi.1000196