2019/10/27 by Anuroop Sriram, Sriram, Anuroop, Jure Žbontar +9 · 2 citations
Engineering · Medicine · #Advanced MRI Techniques and Applications #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #FOS: Electrical engineering #Image and Video Processing (eess.IV) #Medical Imaging Techniques and Applications #Sparse and Compressive Sensing Techniques #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.1910.12325
openalex publication_date 2019/10/27 · openalex created_date 2022/07/28 · openalex updated_date 2026/07/28
Magnetic Resonance Image (MRI) acquisition is an inherently slow process\nwhich has spurred the development of two different acceleration methods:\nacquiring multiple correlated samples simultaneously (parallel imaging) and\nacquiring fewer samples than necessary for traditional signal processing\nmethods (compressed sensing). Both methods provide complementary approaches to\naccelerating the speed of MRI acquisition. In this paper, we present a novel\nmethod to integrate traditional parallel imaging methods into deep neural\nnetworks that is able to generate high quality reconstructions even for high\nacceleration factors. The proposed method, called GrappaNet, performs\nprogressive reconstruction by first mapping the reconstruction problem to a\nsimpler one that can be solved by a traditional parallel imaging methods using\na neural network, followed by an application of a parallel imaging method, and\nfinally fine-tuning the output with another neural network. The entire network\ncan be trained end-to-end. We present experimental results on the recently\nreleased fastMRI dataset and show that GrappaNet can generate higher quality\nreconstructions than competing methods for both 4\× and 8\×\nacceleration.\n