Dynamic MRI reconstruction using nonlocal low-rank constraints based on the deformation corrected signal

Dynamic magnetic resonance imaging (MRI) has a wide application in medical diagnosis. However, because of organ movement, the quality of reconstructed image is not high. Recently, some experiments have shown that MRI reconstruction based on the deformation corrected dynamic signal often achieves better MR image sequence than the reconstruction based on the uncorrected dynamic signal. Moreover, nonlocal low-rank property has ability to preserve the details of the image for MRI reconstruction. Here, we exploit patch-based nonlocal low-rank property which is based on the deformation corrected dynamic signal to achieve better MR images from few measurements. We exploit nonconvex Schatten p-functionals instead of nuclear norm to express the rank approximation. Simulation results indicate that the proposed approach performs better than the previous state-of-the-art algorithms.