A MRI-CT Prostate Registration Using Sparse Representation Technique

In this paper, we propose a new patch-based initial deformation prediction framework for MRI-CT prostate registration. The sparse representation technique is used to estimate the initial deformation between MRI and CT images in a patch-wise fashion. This method can be further refined by many of the existing registration methods after providing a predicted field for initializing prostate MRI-CT registration. Our registration technique was validated with a clinical study of 9 prostate-cancer patients. Our proposed registration was compared with the current intensity-based registration method. This initial investigation of the prostate MRI-CT registration approach using patch-based deformation demonstrated its clinical feasibility as well as its accuracy.