Truncated enhanced constraint for low-rank plus sparse in cardiac dynamic MRI reconstruction

Cardiac dynamic MRI has been widely used in cine and perfusion imaging, in which low-rank and sparse priors play an important role in reconstructing dynamic images with high temporal and spatial resolution from undersampled acquisition. However, the accuracy of singular value estimation may greatly affect the precision of low-rank regularization, resulting in suboptimal reconstruction. A novel reconstruction method employing truncated enhanced constraint for low-rank plus sparse method in cardiac dynamic MRI reconstruction (TRIONES) was proposed to accurately estimate the rank function for dynamic MRI (dMRI) reconstruction. TRIONES uses truncated nuclear norm to assign different shrink values to different singular vectors, preserving the essential image information while effectively eliminating noise. The proposed method was validated in simulation and in vivo experiments for myocardial cine and perfusion MRI reconstruction, comparing with widely used reconstruction methods. The proposed method achieved higher peak signal-to-noise ratio (PSNR), structural similarity (SSIM), and smaller normalized root mean squared error (NRMSE) compared to widely used reconstruction methods in the simulation and in vivo cardiac dynamic MR datasets, suggesting its advantage of accurate rank estimation in cardiac dynamic MRI reconstruction. The proposed TRIONES method can improve low-rank based reconstruction methods for dynamic MRI. It provided more accurate and efficient image information, benefiting the clinical application of heart-related diseases.