Reference-driven MR image reconstruction with sparsity and support constraints

The problem of reconstructing an MR image from limited (and sparsely sampled) k-space data in the presence of a reference image occurs in various applications, including interventional imaging and dynamic contrast-enhanced imaging. This paper addresses the problem using a dictionary composed of three types of basis functions: reference-weighted harmonic functions, wavelets, and pixel/voxel indicator functions. These bases are efficient for representing different image features such as global and local contrast changes from the reference to the target image as well as localized novel image features. The proposed image model and the associated reconstruction algorithm are described. Simulation results are also included to illustrate the improved performance of the proposed method over conventional compressed sensing type reconstruction methods.