Abstract:
|
We develop a model for time-evolving effective connectivity and dynamic changes in causal interactions between many different brain regions. Our approach is a unified framework for reliable and adaptive estimation of state-related changes in effective connectivity, based on switching VAR (SVAR) models. Regimes are uniquely characterized by high dimensional VAR processes, which switch between a finite number of underlying quasi-stationary brain states. The evolution of states and the associated directed dependencies are defined by a Markov chain and the SVAR parameters. The algorithm has three stages: (1.) feature extraction using TV-VAR coefficients; (2.) preliminary regime identification, via clustering of the TV-VAR coefficients; (3.) refined regime segmentation by Kalman smoothing and SVAR parameter estimation via the expectation-maximization (EM) algorithm using the initial estimates from the first two stages. The proposed method was able to identify state-dependent directed connectivity changes via the switching of the VAR states in motor-task fMRI and epileptic seizure EEG data.
|