Spatio-temporal Wavelet Analysis for Functional MRI

• Activity Number: 11
• Type: Topic Contributed
• Date/Time: Sunday, August 8, 2004 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistical Computing
• Author(s): Victor Solo*+ and Christopher Long and Manoach Dara and Emery N. Brown
• Companies: University of Michigan and Massachusetts General Hospital and MGH/NMR Center and Harvard Medical School
• Keywords: MRI ; spatio-temporal ; wavelets

Abstract

Characterizing the spatio-temporal behavior of the BOLD signal in functional MRI (fMRI) is a central issue in understanding brain function. While the nature of functional activation clusters is fundamentally heterogeneous, many current analysis approaches use spatially invariant models that can degrade anatomic boundaries anddistort the underlying functional activation. Furthermore, few analysis approaches use true spatio-temporal continuity in their statistical formulations. To address these issues, we present a novel spatio-temporal wavelet procedure that uses a stimulus con-volved hemodynamic signal plus correlated noise model. The wavelet fits, computed by `1-constrained maximum-likelihood estimation, provide efficient multiscale representations of heterogeneous brain structures, and give well-identified, parsimonious spatial activation estimates that are modulated by the temporal fMRI dynamics. In a study of both simulated data and actual fMRI memory task experiments, our new method gave lower mean-squared error and more focal activation maps than models using standard wavelet and presmoothing techniques.