Neuroimaging research is shifting rapidly toward studying brain dynamism from the perspective of the temporal reconfiguration of functional connectivity, aka dynamic functional connectivity (dFC). However, almost all dFC studies focus only on temporally dynamic properties of the brain, overlooking the importance of within-subject spatial variation. In this work, we will describe and define spatial, temporal, and spatiotemporal dynamics. We also discuss how incorporating space into dFC research can broaden our knowledge of brain function.
We present two analytical techniques that leverage machine learning algorithms to model the brain’s spatial dynamics. A spatially fluid chronnectome model captures moment-to-moment spatial reconfiguration of brain networks at the finest observable scale (voxel-level) and shows how brain networks transiently merge and separate from each other. The second model used the hierarchical models of brain function. Different levels of the hierarchy represent different spatial scales and contain different dynamic information. Spatial dynamic studies have revealed unique patterns of alterations in psychosis, which are hidden and undetectable using previous sFC and dFC techniques. Incorporating space in the dFC analysis also provides a new set of spatial dynamic metrics, inaccessible to previous FC methods, for more accurately capturing the dynamics of brain function.