Whole-brain entrainment and anticipation in fMRI motor task data.

Evidence that the brain adapts its intrinsic rhythms to a task paradigm (entrainment) and that it actively anticipates up-coming tasks has mostly come from analysis of neurophysiological data. Here we show that these phenomena are present in fMRI motor task data as well. We represent brain activity as a series (time resolved) of complex weighted networks and introduce a novel phase clustering algorithm that divides the matrices into functionally meaningful communities with comparable phase content across time and subjects. We find brain-wide patterns of cross-subject synchronizations timed to the task paradigm. This pattern of entrainment features systematic negative deflections prior to task paradigm across many different brain regions. We interpret this pattern as a fMRI-equivalent to the Contingent Negative Variation (CNV) seen in neurophysiological data in anticipation of an up-coming task. A more pronounce version of the CNV-like signal is seen in task specific motor and somatosensory cortex prior to movement, potentially the fMRI-equivalent of the readiness potential (RP). Finally, we show that fMRI data, when represented as complex weighted matrices, full-fills the criteria of structural balance.