Improving Multi-contrast Imaging with Reference Guided Location and Orientation Priors on Edges

Abstract

The multi-contrast magnetic resonance imaging can provide rich clinical and diagnostic information, but it requires long scanning time in data acquisition. In this paper, we propose a reference guided joint reconstruction method to address this problem. The proposed method both incorporates the location and orientation priors on edge regions from a high-resolution reference image into joint sparsity constraints, enabling to effectively reconstruct high-quality multi-contrast images from the under-sampled k-space data. The alternating direction method of multipliers is used to solve the joint sparsity-promoting optimization problem. In addition, a generalized frame with multiple reference images is developed to further improve the reconstruction performance, and the proposed method in combination with parallel imaging is also demonstrated to analyze the feasibility in the practical multi-channel acquisition of multi-contrast images. The experiments have demonstrated the superiority of our proposed method compared to those existing reconstruction technologies in multi-contrast imaging.