Bayesian Spatiotemporal Modeling Using Hierarchical Spatial Priors with Applications to Functional Magnetic Resonance Imaging
Time: 2018-08-22
Published By: Xiaoni Tan
Speaker(s): Galin Jones (University of Minnesota)
Time: 14:00-15:00 August 30, 2018
Venue: Room 77201, Jingchunyuan 78, BICMR
A spatiotemporal Bayesian variable selection model for detecting activation is considered in functional magnetic resonance imaging (fMRI) settings. Following recent research in this area, binary indicator variables for classifying active voxels are used. The spatial dependence in the images is accommodated by applying an areal model to parcels of voxels. The use of parcellation and a spatial hierarchical prior (instead of the popular Ising prior) results in a posterior distribution amenable to exploration with an efficient Markov chain Monte Carlo algorithm. The approach is applied it to simulated data and fMRI data sets.
