Dynamic Cardiac MRI Reconstruction with Weighted Redundant Haar Wavelets

Jun Liu¹, Jeremy Rapin¹, Ti-chiun Chang¹, Alban Lefebvre¹, Michael Zenge², Edgar Mueller², Mariappan S. Nadar¹

High temporal resolution is often desired in Cardiac Magnetic Resonance Imaging (CMRI). Parallel imaging enables the reconstruction with a reduced the number of acquired frequencies, hence accelerating the acquisition, and spiral sampling scheme is widely used in CMRI. In this work, a new approach was proposed for dynamic cardiac MRI reconstruction by applying an L1 regularization based on the weighted 3D redundant Haar wavele for incorporating the smoothness in both spatial and temporal directions. Experiments conducted on a CMRI data with only 15 spokes for each temporal phase demonstrated the effectiveness of the proposed approach.