Assessing the dynamics on functional brain networks using spectral graphy theory

We present an algorithmic pipeline to assess the dynamics on human brain networks based on multimodal resting state functional magnetic resonance imaging (rsfMRI) and diffusion tensor imaging (DTI) data. We employ white matter fiber density information to parcellate the cerebral cortex into functionally homogenous regions, which are used as nodes to construct functional brain networks. Then, the dynamics on the constructed functional networks are assessed using the parameter named propensity for synchronization (PFS) derived from the spectral graph theory. We first demonstrate the ability of PFS in characterizing the dynamics on brain networks by taking the human visual motion perception network (MPN) in resting state and under natural stimulus as test bed systems. The proposed method is then evaluated using the dataset of schizophrenia to demonstrate its application in charactering the abnormalities in functional networks in brain diseases.