End-to-End Variational Networks for Accelerated MRI Reconstruction

By Anuroop Sriram, Jure Zbontar, Tullie Murrell, Aaron Defazio, C. Lawrence Zitnick, Nafissa Yakubova, Florian Knoll, Patricia M. Johnson. International Conference on Medical Image Computing and Computer-Assisted Intervention, 2020

436 citations recorded by the source corpus.

The slow acquisition speed of magnetic resonance imaging (MRI) has led to the development of two complementary methods: acquiring multiple views of the anatomy simultaneously (parallel imaging) and acquiring fewer samples than necessary for traditional signal processing methods (compressed sensing). While the combination of these methods has the potential to allow much faster scan times, reconstruction from such undersampled multi-coil data has remained an open problem. In this paper, we present a new approach to this problem that extends previously proposed variational methods by learning fully end-to-end. Our method obtains new state-of-the-art results on the fastMRI dataset for both brain and knee MRIs.

Subfields: Computational Mechanics, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/6e256528695f5df53ad5a8c3b82ecae0b43f7295

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

Robust Compressed Sensing MRI with Deep Generative Priors (2021, methodology): “Recent research has demonstrated the superior performance of deep neural networks for MR image reconstruction [72, 32, 3, 78, 79].”
Score-based diffusion models for accelerated MRI (2021, methodology): “This can also be seen in Table 2, where we see that our method is on par with the SOTA E2E-varnet on 1D sampling patterns, while significantly outperforming all the methods in 2D sampling patterns.”
Transformers in Medical Imaging: A Survey (2022, result): “It can be seen that ViT model pre-trained on ImageNet produce sharp results as compared to recent CNN-based model [311] and randomly initialized ViT model.”
Results of the 2020 fastMRI Challenge for Machine Learning MR Image Reconstruction (2021, methodology): “Joint-ICNet used a zero-filled reconstruction as the initial estimate and and coil sensitivities were calculated by refining a rough central k-space estimate with a U-Net [14], [41].”