Information theoretic-based similarity measures, in particular mutual information, are widely used for intermodal/intersubject 3D brain image registration. However, conventional mutual information does not take into account spatial dependency between adjacent voxels in images thus reducing its efficacy as a similarity measure in image registration. This paper first presents a review of the existing attempts to incorporate spatial dependency into the computation of mutual information. Then, a recently introduced spatially dependent similarity measure, named spatial mutual information, is extended to 3D brain image registration. This extension also eliminates its artifact for translational misregistration. Finally, the effectiveness of the proposed 3D spatial mutual information as a similarity measure is compared with three existing mutual information measures by applying controlled levels of noise degradation to 3D simulated brain images.
View full article
