CaltechAUTHORS
  A Caltech Library Service

A Bayesian Unobservable/Observable Source Separation Model and Activation Determination in fMRI

Rowe, Daniel B. (2001) A Bayesian Unobservable/Observable Source Separation Model and Activation Determination in fMRI. Social Science Working Paper, 1120R. California Institute of Technology , Pasadena, CA. (Unpublished) https://resolver.caltech.edu/CaltechAUTHORS:20170807-141138268

[img] PDF (sswp 1120R - Jun. 2001) - Submitted Version
See Usage Policy.

1487Kb

Use this Persistent URL to link to this item: https://resolver.caltech.edu/CaltechAUTHORS:20170807-141138268

Abstract

In functional magnetic resonance imaging, the most important question to be answered is that of deciding which statistical method to use in analyzing the data. The statistical analysis is the most crucial task because it determines the statistical activations which are interpreted. In computing the activation level, a standard method is to select assumed to be known reference functions and perform a multiple regression of the time courses on them and a linear trend. In performing the multiple regression, t to F statistics are computed in each voxel then they are colored accordingly. But the most important question is: How do we choose the reference functions? Several different functions have been suggested. This paper, based on Bayesian source separation, determines the underlying source reference functions by instead assuming they are known, assessing a prior distribution for them along with the other parameters and determining them statistically. Both Gibbs sampling and iterated conditional modes algorithms are used to determine marginal posterior mean and joint maximum a posteriori values of the parameters along with statistical activation levels. It was found that the underlying response can be statistically determined and that the iterated conditional modes algorithm performed better than Gibbs sampling.


Item Type:Report or Paper (Working Paper)
Additional Information:Revised version. Original version dated to May 2001.
Group:Social Science Working Papers
Subject Keywords:Bayesian latent variable model, functional magnetic resonance imaging, neuroimaging activation
Series Name:Social Science Working Paper
Issue or Number:1120R
Record Number:CaltechAUTHORS:20170807-141138268
Persistent URL:https://resolver.caltech.edu/CaltechAUTHORS:20170807-141138268
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:79858
Collection:CaltechAUTHORS
Deposited By: Jacquelyn Bussone
Deposited On:07 Aug 2017 22:11
Last Modified:03 Oct 2019 18:24

Repository Staff Only: item control page