Because neuroimaging experiments provide data on multiple brain regions simultaneously, neuroimaging data need not be limited to a single entry in an MTMM matrix. Rather, different brain regions can be sampled to examine the extent to which activity converges or diverges across brain regions. For instance, in considering a measure related to attention, it may be useful to know that task perform ance correlates with activity within the frontal eye field, parietal cortex, and anterior cingulate (all areas involved in attention), but not with activity in the temporal lobe or Broca's area (which is not a component of the system). In this situation, the different brain regions can be equated with different traits in the MTMM matrix. The question becomes, Do anatomically connected or functionally related brain regions (i.e., related traits) show convergence, whereas functionally or anatomically unconnected (i.e., unrelated traits) show divergence? Applying a network approach, one can treat functional couplings (covariance) between regions as separate traits. We can then ask whether different tasks produce convergent or divergent effects on the functional connectivity between regions.
Considered in this framework, neuroimaging is highly compatible with the MTMM approach, allowing the assessment of convergence and divergence across stimulation methods and the brain regions activated by those methods. However, this approach is rarely formally applied in the neuroimaging field. This in part reflects the difficulty in ascribing brain activations in a given region to a specific function. For instance, although the dorsolateral prefrontal cortex frequently activates during working memory tasks, it also activates during tasks that are not specifically related to working memory (D'Esposito, Ballard, Aguirre, & Zarahn, 1998). Indeed, the multitude of functions proposed for the prefrontal cortex makes it unlikely that a single discrete process can explain all the varied tasks that lead to increased activity in the region (Duncan & Owen, 2000). Thus, it would be unwise to assume that activation of the dorsolateral prefrontal cortex (or other brain regions involved in working memory tasks) necessarily indicate the involvement of working memory in a given task. On the other hand, if we have three areas, each of which are engaged by multiple cognitive tasks, but that only show simultaneous activation during working memory, then the multiregion approach could prove highly useful.
Was this article helpful?