Skip to Main Content
In functional magnetic resonance imaging (fMRI) data, activated voxels are usually very small in number and are embedded in a mass of inactive voxels. For clustering analysis, this situation generates an ill-balanced data problem among different classes of voxels. In this paper we propose a novel method to overcome the ill-balanced data problem by eliminating the inactive voxels prior to the application of the clustering algorithm. The proposed method is based on a region homogeneity criterion in which the homogeneity is determined on the basis of the variance explained by the first principal component in the principal component analysis (PCA) of a given voxel and its neighbors. We present the results of application of the proposed method for both simulated and real fMRI data. The results advocate that the proposed method effectively solves the ill-balanced data problem by eliminating most of the inactive voxels while retaining the functionally active voxels.