Between-subject variance in resting-state fMRI connectivity predicts fMRI activation in a language task

Abstract

The spatial response observed during fMRI experiments is generally highly variable across subjects. In the current study, we disentangled the inter-subject variance that reflects differences in the network topology of the brain and the variance caused by other effects (e.g. task compliance or engagement). Using resting-state connectivity and task-based activation during a language task, we found signif­icant correlations between variance across the two measures. The strength with which a region is connected to the language network during rest is found to be predictive of how active it is during language processing. This finding demonstrates that a significant portion of the inter-subject vari­ability observed in fMRI studies can be related to inter-individual differences in brain organization.