Methods for Detecting Functional Connectivity Change Points in fMRI Data

• Activity Number: 215
• Type: Invited
• Date/Time: Monday, August 5, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Imaging
• Author(s): Ivor Cribben*+ and Tor D. Wager and Martin Lindquist
• Companies: University of Alberta School of Business and University of Colorado, Boulder and Johns Hopkins Bloomberg School of Public Health
• Keywords: network change points ; graph based change point detection ; graphical lasso ; dynamic networks ; stability selection ; stationary bootstrap

Abstract

Recently in functional magnetic resonance imaging (fMRI) studies there has been an increased interest in understanding the dynamic manner in which brain regions communicate with one another, as subjects perform a set of experimental tasks or as their psychological state changes. In this work, we extend Dynamic Connectivity Regression (DCR), a technique used for detecting temporal change points in functional connectivity (or network structure) between brain regions where the number and location of the change points are unknown, firstly by introducing a new algorithm for detecting the changes in functional connectivity for single-subject fMRI data. Secondly, we propose a new DCR technique that utilizes Bayesian network modeling. Lastly, we discuss methods that combine information across subjects for multi-subject fMRI studies as variability between subjects makes the use of these data sets challenging. The new methods are applied to various simulated data sets as well as to fMRI data sets. The results illustrate the method's ability to observe how the networks between different brain regions change as the emotional state of subjects change.