|
Abstract:
|
We propose a flexible approach for modeling and spatially clustering voxel-wise functional response curves for multi-subject fMRI data. Our goal is to cluster the brain into regions with similar response curves over levels of a stimulus, and to estimate these region-wide curves and their variability at the level of subjects and voxels. We apply functional data analytic modeling techniques to response functions at each voxel, and employ a model-based unsupervised spatial clustering algorithm to estimate regions with homogeneous response profiles. Curves are modeled using a polynomial spline basis, and are registered by subject. Spline coefficients are defined as parameters in a random-effects model, allowing us to borrow strength across subjects and voxels, and are decomposed into a regional average and a random effect component.
|
