|
Activity Number:
|
204
- Bayesian Methods for the Analysis of Complex Brain Imaging Data
|
|
Type:
|
Topic-Contributed
|
|
Date/Time:
|
Tuesday, August 10, 2021 : 1:30 PM to 3:20 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
|
Abstract #317063
|
|
|
Title:
|
Efficient Estimation of Brain Activation with Cortical Surface Data Using EM with a Bayesian General Linear Model
|
|
Author(s):
|
Daniel Spencer* and Amanda Mejia
|
|
Companies:
|
Indiana University and Indiana University
|
|
Keywords:
|
fMRI studies;
expectation maximization;
SPDE prior;
brain activation;
general linear model
|
|
Abstract:
|
Functional magnetic resonance imaging (fMRI) data require robust analysis methods to provide meaning inference on low-signal data. Bayesian methods in particular allow for the inclusion of expected behavior from prior study into an analysis, increasing the power of the results while circumventing problems that arise in classical analyses, including the effects of smoothing results and sensitivity to multiple comparison testing corrections. Recent work by Mejia et al. (2020) developed a Bayesian general linear model for cortical surface fMRI (cs-fMRI) data reliant on stochastic partial differential equation (SPDE) priors which relies on the computational efficiencies of the integrated nested Laplace approximation (INLA) to perform this powerful analysis. In this article we develop an exact Bayesian analysis method for the GLM, employing an expectation-maximization (EM) algorithm to find maximum a posteriori (MAP) estimates of task-based regressors on cs-fMRI data. Our proposed method is compared to the Bayesian GLM developed by Mejia et al. (2020), as well as a classical GLM on simulated data. A validation of the method on data from the Human Connectome Project is also provided.
|
Authors who are presenting talks have a * after their name.
