Abstract #0540
Modeling of BOLD Components in the Statistical Analysis of Perfusion-Based fMRI Experiments
Roller E, Liu T, Restom K
UCSD, UCSD Center for fMRI
Perfusion based fMRI with arterial spin labeling (ASL) suffers from inherently low sensitivity. As a result, optimal analysis methods are required to make ASL a robust and widely applicable tool. Here we present an extension of a general linear model for the analysis of ASL data. We show that the incorporation of a BOLD-weighted static tissue term is critical for achieving optimal statistical power for acquisitions with long echo times (TE= 30 ms) but can be neglected for acquisitions with short echo times (TE = 9 ms).
