Magnetic resonance parameter mapping using model‐guided self‐supervised deep learning

Liu, Fang; Kijowski, Richard; Fakhri, Georges El; Feng, Li

doi:10.1002/mrm.28659

Cited by 61 publications

(57 citation statements)

References 54 publications

(152 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this setting, physics information has primarily been incorporated to the loss function during training, similar to Eqs. ( 29) and (30) [9,60,61]. Recently, networks that reflect low-rank subspace constraints have been proposed [62].…”

Section: Inverse Problems In Mri With Non-linear Forward Modelsmentioning

confidence: 99%

Physics-Driven Deep Learning for Computational Magnetic Resonance Imaging

Hammernik¹,

Küstner²,

Huang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Physics-driven deep learning methods have emerged as a powerful tool for computational magnetic resonance imaging (MRI) problems, pushing reconstruction performance to new limits. This article provides an overview of the recent developments in incorporating physics information into learningbased MRI reconstruction. We consider inverse problems with both linear and non-linear forward models for computational MRI, and review the classical approaches for solving these. We then focus on physics-driven deep learning approaches, covering physics-driven loss functions, plug-and-play methods, generative models, and unrolled networks. We highlight domain-specific challenges such as real-and complex-valued building blocks of neural networks, and translational applications in MRI with linear and non-linear forward models. Finally, we discuss common issues and open challenges, and draw connections to the importance of physics-driven learning when combined with other downstream tasks in the medical imaging pipeline.

show abstract

Section: Inverse Problems In Mri With Non-linear Forward Modelsmentioning

confidence: 99%

Physics-Driven Deep Learning for Computational Magnetic Resonance Imaging

Hammernik¹,

Küstner²,

Huang³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…In turn, heavy demand for training data limits applicability of deep reconstruction since collection of large MRI datasets at a single site is challenging [38], [39], and cross-site transfer of imaging data raises patient privacy concerns [21]. To facilitate dataset curation, unpaired [13], [40], [41], self-supervised [42]- [45], or transfer [6], [33] learning strategies were previously proposed. However, these methods still require centralized model training following accumulation of a sufficiently diverse dataset.…”

Section: Contributionsmentioning

confidence: 99%

Federated Learning of Generative Image Priors for MRI Reconstruction

Elmas¹,

Dar²,

Korkmaz³

et al. 2022

Preprint

View full text Add to dashboard Cite

Multi-institutional efforts can facilitate training of deep MRI reconstruction models, albeit privacy risks arise during cross-site sharing of imaging data. Federated learning (FL) has recently been introduced to address privacy concerns by enabling distributed training without transfer of imaging data. Existing FL methods for MRI reconstruction employ conditional models to map from undersampled to fully-sampled acquisitions via explicit knowledge of the imaging operator. Since conditional models generalize poorly across different acceleration rates or sampling densities, imaging operators must be fixed between training and testing, and they are typically matched across sites. To improve generalization and flexibility in multi-institutional collaborations, here we introduce a novel method for MRI reconstruction based on Federated learning of Generative IMage Priors (FedGIMP). FedGIMP leverages a two-stage approach: cross-site learning of a generative MRI prior, and subject-specific injection of the imaging operator. The global MRI prior is learned via an unconditional adversarial model that synthesizes high-quality MR images based on latent variables. Specificity in the prior is preserved via a mapper subnetwork that produces sitespecific latents. During inference, the prior is combined with subject-specific imaging operators to enable reconstruction, and further adapted to individual test samples by minimizing data-consistency loss. Comprehensive experiments on multi-institutional datasets clearly demonstrate enhanced generalization performance of FedGIMP against site-specific and federated methods based on conditional models, as well as traditional reconstruction methods.

show abstract

“…Neural network architectures have gained popularity in the literature [12], [13] due to their ability to achieve regularization by learning the transform parameters without the need for hand-crafting the regularization function. Neural networks offer flexibility to determine the spatio-temporal redundancy in the dynamic MR images [14], [15].…”

Section: Introductionmentioning

confidence: 99%