Analyzing functional magnetic resonance imaging (fMRI) data has revealed localized brain regions and functional networks that exhibit consistent properties across subjects. Moreover, alterations in such properties are often associated with mental illnesses, including major depressive disorder (MDD). Separate fMRI analyses are typically conducted to address objectives concerning localized brain activity patterns and functional networks (connectivity). A key statistical challenge is to develop methodology to address both of these objectives within a unified modeling framework. We present a Bayesian spatial model, extending work by Bowman et al., 2008, that yields inferences for both localized activation patterns and network properties. The extended model is applicable to both task and resting-state fMRI data. We perform estimation using MCMC methods implemented via Gibbs sampler. We apply our Bayesian model to data from an fMRI study of MDD.
Keywords: Spatial modeling; Brain imaging; Bayesian modeling
Biography: DuBois Bowman is an Associate Professor of Biostatistics and Bioinformatics at Emory University and Director of the Center for Biomedical Imaging Statistics (CBIS) at Emory. He has built a program of research involving the development of statistical methods for brain imaging data, including fMRI, PET, and DTI. His methodological research interests focus on 1) spatio-temporal modeling of functional neuroimaging data, 2) characterizing functional connectivity in the brain, and 3) prediction, e.g. predicting brain responses to psychiatric treatments.