JSM 2004 - Toronto

Abstract #301677

This is the preliminary program for the 2004 Joint Statistical Meetings in Toronto, Canada. Currently included in this program is the "technical" program, schedule of invited, topic contributed, regular contributed and poster sessions; Continuing Education courses (August 7-10, 2004); and Committee and Business Meetings. This on-line program will be updated frequently to reflect the most current revisions.

To View the Program:
You may choose to view all activities of the program or just parts of it at any one time. All activities are arranged by date and time.

The views expressed here are those of the individual authors
and not necessarily those of the ASA or its board, officers, or staff.

Back to main JSM 2004 Program page
Activity Number: 78
Type: Topic Contributed
Date/Time: Monday, August 9, 2004 : 8:30 AM to 10:20 AM
Sponsor: Biometrics Section
Abstract - #301677
Title: Autocorrelation Model Selection in fMRI Analysis
Author(s): Jeanette Mumford*+ and Thomas E. Nichols and Wen-Lin Luo
Companies: University of Michigan and University of Michigan and University of Michigan
Address: Department of Biostatistics, Ann Arbor, MI, 48105,
Keywords: fMRI time series ; correlation model identification ; goodness of fit
Abstract:

A functional magnetic resonance imaging (fMRI) data exhibit temporal autocorrelation. It is crucial to identify an accurate autocorrelation model so that efficient inferences can be obtained. In time series analysis, ACF and PACF are usually used to explore the possible autocorrelation model. In neuroimaging, this work breaks down since it is not practical to look at 100,000 plots, one for each volume element. Therefore we need a way to efficiently identify the best autocorrelation model over the entire brain. We apply model selection techniques to choose the fMRI autocorrelation model. We use simulation to understand the performance of different model selection criteria (AIC, BIC, MDL, ICOMP, and NURE) in selecting the autocorrelation model. Based on the simulation results and combined with the graphical exploration in time and frequency domain, we develop a systematic approach to modeling fMRI autocorrelation. With our previous work on linear model diagnosis, we can assess the goodness of fit of both the mean and autocorrelation model. We demonstrate our method by a real fMRI data analysis.

  • The address information is for the authors that have a + after their name.
  • Authors who are presenting talks have a * after their name.

Back to the full JSM 2004 program

JSM 2004 For information, contact jsm@amstat.org or phone (888) 231-3473. If you have questions about the Continuing Education program, please contact the Education Department.
Revised March 2004