MR Image Reconstruction via Sparse Representation: Modeling and Algorithm.

摘要: To reduce acquisition time in magnetic resonance (MR) imaging, compressive sensing and sparse representation techniques have been developed to reconstruct MR images with partially acquired data. Although this has a hot research topic the field, it not used clinically due three inherent problems of its current framework: potential loss fine structures, difficulty predefine model parameters, long reconstruction time. The aim work is tackle these problems. We propose minimize total variation underlying image, together `1 norm coefficients using trained dictionary, as well fidelity term. Using dictionary can take advantage prior knowledge hence improve accuracy reconstruction. Our data constraint derived from likelihood estimator recovering error partial k-space robustness parameter selection. Moreover, simple efficient numerical scheme provided solve faster. consequent experiments on both synthetic vivo indicate improvement proposed preserving structure, reducing computational cost, flexibility decision.