Raquel Prado of the Department of Applied Mathematics and Statistics at the University of California Santa Cruz. The seminar will be held at3:30pm, in 115 Ford Hall, May 5, 2016.
Title: Bayesian approaches for brain activation and connectivity
Abstract: We consider Bayesian models for jointly inferring activation and connectivity from brain signals. We present models for fMRI data from multiple regions of interest (ROIs) that allow us to obtain simultaneous estimation of connectivity networks and hemodynamic response functions that are region, task, and subject-specific, while taking into account variations across subjects and experimental conditions. We illustrate these approaches through the analysis of various simulated datasets and a human dataset from a stroke study that involved multiple subjects. We also present computationally feasible Bayesian approaches for detecting activation in complex-valued fMRI data. We apply these models to the analysis of experimentally realistic synthetic complex-valued fMRI as well as human complex-valued fMRI.