Deep Generative Adversarial Networks for Compressed Sensing Automates MRI

Mardani, Morteza; Gong, Enhao; Cheng, Joseph Y.; Vasanawala, Shreyas; Zaharchuk, Greg; Alley, Marcus T.; Thakur, Neil; Han, Song; Dally, William J.; Pauly, John M.; Liu, Xing

doi:10.48550/arxiv.1706.00051

Cited by 59 publications

(59 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…GANs ability to provide realistic, high texture quality images was exploited in [96] for CS MR image reconstruction. A deep residual network with skip connections is trained as the generator that learns to remove the aliasing artifacts by projecting it onto a low-dimensional manifold containing the desired, high-quality data.…”

Section: Residual Networkmentioning

confidence: 99%

“…Another category of methods uses deep networks to mitigate noise and aliasing artifacts in the MRI reconstruction. SDAE [92], CNNs [139], residual networks [50,77,109,131], or GANs [96,156] were successfully validated as suitable architectures for CS-DL MRI and have all shown great potential, outperforming conventional CS techniques for undersampled MRI.…”

Section: Lessons From Domain-specific Cs-dlmentioning

confidence: 99%

See 1 more Smart Citation

Deep Learning Techniques for Compressive Sensing-Based Reconstruction and Inference -- A Ubiquitous Systems Perspective

Machidon,

Pejovic

2021

Preprint

View full text Add to dashboard Cite

Compressive sensing (CS) is a mathematically elegant tool for reducing the sampling rate, potentially bringing contextawareness to a wider range of devices. Nevertheless, practical issues with the sampling and reconstruction algorithms prevent further proliferation of CS in real world domains, especially among heterogeneous ubiquitous devices. Deep learning (DL) naturally complements CS for adapting the sampling matrix, reconstructing the signal, and learning form the compressed samples. While the CS-DL integration has received substantial research interest recently, it has not yet been thoroughly surveyed, nor has the light been shed on practical issues towards bringing the CS-DL to real world implementations in the ubicomp domain. In this paper we identify main possible ways in which CS and DL can interplay, extract key ideas for making CS-DL efficient, identify major trends in CS-DL research space, and derive guidelines for future evolution of CS-DL within the ubicomp domain. CCS Concepts: • Computer systems organization → Sensor networks; • Human-centered computing → Ubiquitous computing; • Computing methodologies → Neural networks.

show abstract

Section: Residual Networkmentioning

confidence: 99%

Section: Lessons From Domain-specific Cs-dlmentioning

confidence: 99%

Deep Learning Techniques for Compressive Sensing-Based Reconstruction and Inference -- A Ubiquitous Systems Perspective

Machidon,

Pejovic

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…With the rise of machine learning, many methods (Zhu et al, 2018;Mardani et al, 2017;Shen et al, 2019;Würfl et al, 2018;Ghani & Karl, 2018;Wei et al, 2020) have been proposed for medical image reconstruction using a small number of measurements. Most of these methods are supervised learning techniques.…”

Section: Introductionmentioning

confidence: 99%

Solving Inverse Problems in Medical Imaging with Score-Based Generative Models

Song¹,

Shen²,

Liu³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Reconstructing medical images from partial measurements is an important inverse problem in Computed Tomography (CT) and Magnetic Resonance Imaging (MRI). Existing solutions based on machine learning typically train a model to directly map measurements to medical images, leveraging a training dataset of paired images and measurements. These measurements are typically synthesized from images using a fixed physical model of the measurement process, which hinders the generalization capability of models to unknown measurement processes. To address this issue, we propose a fully unsupervised technique for inverse problem solving, leveraging the recently introduced score-based generative models. Specifically, we first train a score-based generative model on medical images to capture their prior distribution. Given measurements and a physical model of the measurement process at test time, we introduce a sampling method to reconstruct an image consistent with both the prior and the observed measurements. Our method does not assume a fixed measurement process during training, and can thus be flexibly adapted to different measurement processes at test time. Empirically, we observe comparable or better performance to supervised learning techniques in several medical imaging tasks in CT and MRI, while demonstrating significantly better generalization to unknown measurement processes.

show abstract

“…After this prelude, the mathematical representation of the DCS is given as y = ΦG θ (z), where G θ is the deep learning based R k → R n generator. One can find utilization of generative models in DCS and how they generalize beyond classical CS models in Wu et al [2019], Mardani et al [2017], Bora et al [2017], Mardani et al [2018], Sun et al [2020], Van Veen et al [2018]. There exists certain works which address the robust deep compressed sensing for defending against adversarial attack.…”

Section: Introduction and Related Workmentioning

confidence: 99%

“…Further elaborations are provided through the paper body. Several applications rely on supremacy of DCS models including medical resonance imaging (MRI) Mardani et al [2017], , Quan et al [2018], Yi et al [2019], Jiang et al [2019], Qiusheng et al [2020], Lee et al [2017] and wireless neural recording Sun et al [2016].…”

Section: Introduction and Related Workmentioning

confidence: 99%

Generative Model Adversarial Training for Deep Compressed Sensing

Esmaeili

2021

Preprint

View full text Add to dashboard Cite

Deep compressed sensing assumes the data has sparse representation in a latent space, i.e., it is intrinsically of low-dimension. The original data is assumed to be mapped from a low-dimensional space through a low-to-highdimensional generator. In this work, we propound how to design such a low-to-high dimensional deep learning-based generator suiting for compressed sensing, while satisfying robustness to universal adversarial perturbations in the latent domain. We also justify why the noise is considered in the latent space. The work is also buttressed with theoretical analysis on the robustness of the trained generator to adversarial perturbations. Experiments on real-world datasets are provided to substantiate the efficacy of the proposed generative model adversarial training for deep compressed sensing.

show abstract

Deep Generative Adversarial Networks for Compressed Sensing Automates MRI

Cited by 59 publications

References 11 publications

Deep Learning Techniques for Compressive Sensing-Based Reconstruction and Inference -- A Ubiquitous Systems Perspective

Deep Learning Techniques for Compressive Sensing-Based Reconstruction and Inference -- A Ubiquitous Systems Perspective

Solving Inverse Problems in Medical Imaging with Score-Based Generative Models

Generative Model Adversarial Training for Deep Compressed Sensing

Contact Info

Product

Resources

About