Identification of State Changes from Spontaneous Fluctuations in FMRI Data

Meghan Robinson¹, Javier Gonzalez-Castillo¹, Souheil Inati², Daniel Handwerker¹, Peter Bandettini¹, ²

The brain is continuously adapting to respond to the environment. In the course of a rest scan, subjects are expected to engage in different cognitive states for variable periods of time (i.e., relaxing, planning dinner, napping, etc.). Each of these states brings the brain into a different pattern of overall connectivity. Still, some level of stability is expected for the periods when subjects remain in a given cognitive state. Here, we propose a way to automatically identify such cognitive states looking at how the fMRI signal from different cortical regions pass in-and-out of synchrony with each other as time progresses.