Bayesian Latent Variable Models for MR Imaging Data with Multiple Outcomes

• Activity Number: 493
• Type: Contributed
• Date/Time: Wednesday, August 7, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Imaging
• Author(s): Xiao Wu*+ and Michael Daniels
• Companies: University of Florida and The University of Texas at Austin
• Keywords: latent variable ; Bayesian ; multiple outcomes ; autocorrelation

Abstract

MR imaging analysis involves combining multiple measures of interest over time. In this research, we describe a Bayesian approach to model longitudinal multidimensional continuous outcomes with latent variables and partial autocorrelation selection priors. Dependence among different outcomes is induced through latent variables and covariance matrices are estimated simultaneously by using nonparametric priors. A Markov chain Monte Carlo algorithm is proposed for estimating the posterior distributions of the parameters and latent variables. This method is illustrated using data from a Duchenne Muscular Dystrophy study on changes in muscle imaging data to capture disease progression over time.