A Wavelet-based Multiresolution Medical Image Registration Strategy Combining Mutual Information with Spatial Information

Abstract

Abstract. Mutual information is a widely-used metric in medical image registration, which achieves good results; however it can lead to misregistration because of the absence of spatial information. One of the works to overcome this problem is proposed by [10] in which a gradient-based term is combined with mutual information. In this paper, we improve the method in [10] and propose a registration strategy which is more suitable when multiresolution scheme is adopted. Different from [10], the calculations of both the images with different resolutions and their corresponding gradient-like images are integrated under the wavelet transform. Mutual information (MI) and spatial information (SI) are respectively calculated from the low-frequency coefficients and high-frequency coefficients of wavelet transform on each resolution layer; and a hybrid metric is adopted to combine MI and SI together. The proposed registration strategy can run faster than the method in [10] when multiresolution scheme is used. Additionally, compared to the other existing works where wavelet transform is adopted for registration, we apply wavelet transform to the problem of 3D image registration and make use of both of the high frequency coefficients and low frequency coefficients. Experiments show that the proposed method behaves well in MR-MR and CT-MR registration problems.