A Nonrigid Registration Framework Using Spatially Encoded Mutual Information and Free-Form Deformations

Xiahai Zhuang,Simon Arridge,David J. Hawkes,Sebastien Ourselin

DOI: https://doi.org/10.1109/tmi.2011.2150240

IF: 10.6

2011-01-01

IEEE Transactions on Medical Imaging

Abstract:Mutual information (MI) registration including spatial information has been shown to perform better than the traditional MI measures for certain nonrigid registration tasks. In this work, we first provide new insight to problems of the MI-based registration and propose to use the spatially encoded mutual information (SEMI) to tackle these problems. To encode spatial information, we propose a hierarchical weighting scheme to differentiate the contribution of sample points to a set of entropy measures, which are associated to spatial variable values. By using free-form deformations (FFDs) as the transformation model, we can first define the spatial variable using the set of FFD control points, and then propose a local ascent optimization scheme for nonrigid SEMI registration. The proposed SEMI registration can improve the registration accuracy in the nonrigid cases where the traditional MI is challenged due to intensity distortion, contrast enhancement, or different imaging modalities. It also has a similar computation complexity to the registration using traditional MI measures, improving up to two orders of magnitude of computation time compared to the traditional schemes. We validate our algorithms using phantom brain MRI, simulated dynamic contrast enhanced mangetic resonance imaging (MRI) of the liver, and in vivo cardiac MRI. The results show that the SEMI registration significantly outperforms the traditional MI registration.