A1171
Title: A predictor-informed multi-subject Bayesian approach for dynamic functional connectivity
Authors: Michele Guindani - University of California Los Angeles (United States) [presenting]
Abstract: Time-Varying Functional Connectivity (TVFC) investigates dynamic interactions between brain regions over fMRI experiments. These changes can be modulated by underlying physiological mechanisms such as attention or cognitive effort. A multi-subject Bayesian framework is proposed to estimate dynamic functional networks based on time-varying exogenous physiological covariates. A non-homogeneous hidden Markov model is used to classify fMRI time series into latent neurological states, allowing for the estimation of recurrent connectivity patterns and the sharing of networks among subjects. The model also assumes sparsity in network structures via shrinkage priors, with edge selection achieved through a multi-comparison procedure for shrinkage-based inferences with Bayesian false discovery rate control. The framework is applied to a resting-state experiment with concurrent pupillometry measurements, revealing the effects of changes in pupil dilation on the subjects' propensity to change connectivity states.
