Brain state change detection via fiber-centered functional connectivity analysis

Structural and functional brain connectivity has been extensively studied via diffusion tensor imaging (DTI) and functional MRI (fMRI) in recent years. An important aspect that has not been adequately addressed before is the connectivity state change in structurally-connected brain regions. In this paper, we present an intuitive approach that extracts feature vectors describing the functional connectivity state of the brain with the guidance of DTI data. The general idea is that the functional connectivity patterns of all of the fiber-connected voxels within the brain are concatenated into a feature vector to represent the brain's state, and brain state change points are determined by the abrupt changes of the vector patterns calculated by the sliding window approach. Our results show that we can detect meaningful critical brain state change time points in task-based fMRI and natural stimulus fMRI data. In particular, the detected brain state change points in task-based fMRI data well corresponded to the stimulus task paradigm given to the subjects, providing validation to the proposed brain state change detection approach.