CANCELLED: Defense: Bayesian modeling of complex-valued fMRI signals

Speaker Name: 
Cheng-Han Yu
Speaker Title: 
PhD Candidate
Speaker Organization: 
Statistics & Applied Mathematics
Start Time: 
Tuesday, March 13, 2018 - 9:45am
Raquel Prado

Abstract:  Detecting which voxels or brain regions are activated by an external stimulus is a common objective in functional magnetic resonance (fMRI) studies, however, most studies use magnitude-only fMRI data and discard the phase data. We consider a set of statistical models for detecting brain activation at the voxel level that make use of the entire complex-valued fMRI time courses. Using a complex normal spike-and-slab mixture prior, we develop a general complex-valued expectation-maximization variable selection (C-EMVS) algorithm with autoregressive structure that efficiently and better accurately detects fMRI brain activation at the lowest voxel level. A corresponding complex-valued stochastic search Markov chain Monte Carlo algorithm is also developed to quantify the uncertainty of the activation strength of each voxel in the image.

To further improve detection performance, a computationally efficient Bayesian spatial model is developed to explicitly model the spatial dependence across voxels through kernel convolution. The model encourages voxels to be activated in clusters and is able to eliminate isolated voxels that are incorrectly labeled as active in models that do not assume a spatial structure.

Another key topic in neuroscience studies is brain connectivity or how brain regions interact with each other. We generalize the spatial model mentioned above from a single-subject to a multi-subject model, and use the information from multiple subjects to infer brain functional connectivity. This model is general and practical due to its ability to infer brain activation and connectivity simultaneously.

We illustrate the performance of our statistical models and tools in extensive and physically realistic simulation studies and also in the analysis of human complex-valued fMRI data from a task-related study.