A neighborhood method for statistical analysis of fMRI data

Abstract

In an effort to cope with the fact that functional magnetic resonance imaging (fMRI) data are spatio-temporally correlated, we propose a novel statistical method with a view to improve the detection of brain regions with increased neuronal activity in fMRI. In this method, we make use of information from neighboring voxels of a voxel, for estimation at the voxel. We examined performance of the method against the statistical parametric mapping (SPM) method using both simulated and real data. The proposed method is shown to be considerably better than the SPM in the context of Receiver Operating Characteristics (ROC) curves.