|
Activity Number:
|
455
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #306297 |
|
Title:
|
Wild Bootstrap for Functional Magnetic Resonance Imaging Data
|
|
Author(s):
|
Hongtu Zhu*+ and Bradley S. Peterson
|
|
Companies:
|
Columbia University and New York State Psychiatric Institute and Columbia University and New York State Psychiatric Institute
|
|
Address:
|
Department of Child Psychiatry, New york, NY, 10032,
|
|
Keywords:
|
hypothesis test ; functional MRI ; multiple testing ; statistical parametric mapping
|
|
Abstract:
|
Multiple testing problems arise commonly in the analysis of functional magnetic resonance imaging (fMRI) data. Much effort has been devoted to developing multiple testing procedures based on the familywise error rate to control overall type I errors. For instance, statistical methods---including random field theory and permutation method---have been used widely to calculate corrected p-values, accounting for tests. In this paper, we discuss wild bootstrap method and consider its application in fMRI data. The finite sample performance of wild bootstrap method is investigated with Monte Carlo experiments. The simulation results suggest the wild bootstrap is a good approach under certain conditions compared to other approaches (e.g., permutation). The results also are illustrated by applications to two real databases.
|
