Modeling Functional Connectivity in the Human Brain with Incorporation of Structural Connectivity

• Activity Number: 33
• Type: Contributed
• Date/Time: Sunday, July 29, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistical Learning and Data Mining
• Author(s): Wenqiong Xue*+ and F. DuBois Bowman
• Companies: Emory University and Emory University
• Keywords: Functional Connectivity ; Structural connectivity ; Bayesian analysis ; MCMC ; fMRI/DTI ; Neuroimaging

Abstract

Recent innovations in neuroimaging technology provide opportunities for researchers to investigate the anatomy as well as the function of the human brain. Existing approaches for connectivity generally examine functional properties of resting-state or task-related connectivity between brain regions or separately examine structural properties. We present a unified Bayesian framework to determine functional connectivity utilizing supplementary knowledge of associated structural connections. Our functional connectivity measure assesses functional coherence, reflecting simultaneously elevated regional brain activity identified from fMRI data. Our structural connectivity information is drawn from DTI data, which we use to quantify probabilities of structural connectivity between regions. We assess the functional hierarchy of coherent region pairs by measuring the ascendancy, which compares the marginal probabilities of elevated activity between regions. We demonstrate the use of our model using data from an auditory processing study. We further illustrate the advantages of our method by conducting simulation studies to compare it to methods that only incorporate functional information.