Deep Generative Adversarial Neural Networks for Compressive Sensing (GANCS) MRI

By Morteza Mardani, Enhao Gong, J. Cheng, Shreyas S. Vasanawala, Greg Zaharchuk, Lei Xing, John M. Pauly. 2018

552 citations recorded by the source corpus.

Undersampled magnetic resonance image (MRI) reconstruction is typically an ill-posed linear inverse task. The time and resource intensive computations require tradeoffs between <italic>accuracy</italic> and <italic>speed</italic>. In addition, state-of-the-art compressed sensing (CS) analytics are not cognizant of the image <italic>diagnostic quality</italic>. To address these challenges, we propose a novel CS framework that uses generative adversarial networks (GAN) to model the (low-dimensional) manifold of high-quality MR images. Leveraging a mixture of least-squares (LS) GANs and pixel-wise <inline-formula> <tex-math notation="LaTeX">$\ell _{1}/\ell _{2}$ </tex-math></inline-formula> cost, a deep residual network with skip connections is trained as the generator that learns to remove the <italic>aliasing</italic> artifacts by projecting onto the image manifold. The LSGAN learns the texture details, while the <inline-formula> <tex-math notation="LaTeX">$\ell _{1}/\ell _{2}$ </tex-math></inline-formula> cost suppresses high-frequency noise. A discriminator network, which is a multilayer convolutional neural network (CNN), plays the role of a perceptual cost that is then jointly trained based on high-quality MR images to score the quality of retrieved images. In the operational phase, an initial aliased estimate (e.g., simply obtained by zero-filling) is propagated into the trained generator to output the desired reconstruction. This demands a very low computational overhead. Extensive evaluations are performed on a large contrast-enhanced MR dataset of pediatric patients. Images rated by expert radiologists corroborate that GANCS retrieves higher quality images with improved fine texture details compared with conventional Wavelet-based and dictionary-learning-based CS schemes as well as with deep-learning-based schemes using pixel-wise training. In addition, it offers reconstruction times of under a few milliseconds, which are two orders of magnitude faster than the current state-of-the-art CS-MRI schemes.

Subfields: Computational Mechanics, Radiation, Radiology, Nuclear Medicine and Imaging

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 Radiation

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

An overview of deep learning in medical imaging focusing on MRI (2018, background): “Employing least-squares generative adversarial networks (GANs) that learns texture details and suppresses high-frequency noise, [138] created a novel compressed sensing framework that can produce diagnostic quality reconstructions “on the fly” (30 ms).”
FastSurfer - A fast and accurate deep learning based neuroimaging pipeline (2019, unknown): “…al., 2017; Varghese et al., 2017; Rezaei et al., 2017; Roa-Barco et al., 2017; Chen and Konukoglu, 2018), MR image reconstruction (Jin et al., 2017; Mardani et al., 2019; Schlemper et al., 2018; Yang et al., 2018; Dedmari et al., 2018), prediction of brain related diseases and their progression…”
Monarch: Expressive Structured Matrices for Efficient and Accurate Training (2022, methodology): “To reduce dependence on labeled data, unsupervised methods have used generative adversarial networks (Cole et al., 2020; Mardani et al., 2018), self-supervised learning (Yaman et al., 2020), dictionary learning (Lahiri et al., 2021), and untrained networks (Darestani & Heckel, 2021).”
Deep Magnetic Resonance Image Reconstruction: Inverse Problems Meet Neural Networks (2020, unknown): “Due to its outstanding performance in image-to-image translation, GAN has been exploited in MRI reconstruction [16], [17] to correct the aliasing artifacts in the zero-filled reconstruction from undersampled MR data.”
GAN-Based Projector for Faster Recovery With Convergence Guarantees in Linear Inverse Problems (2019, background): “However, in [24], the operator A is integrated into training the GAN, limiting it to a particular inverse problem.”
NeRP: Implicit Neural Representation Learning with Prior Embedding for Sparsely Sampled Image Reconstruction (2021, methodology): “Later, adversarial learning was introduced to train the CNN model for obtaining better reconstruction image quality [10], [11].”
Deep-Learning Methods for Parallel Magnetic Resonance Imaging Reconstruction: A Survey of the Current Approaches, Trends, and Issues (2019, methodology): “Comparable approaches were used in the context of MRI reconstruction [73–78].”
Intermediate Layer Optimization for Inverse Problems using Deep Generative Models (2021, background): “On the empirical side, a central open problem would be the application of our framework in other domains like medical imaging, but that would require pre-trained generative models e.g. for high-resolution MRI images.”
ADMM-CSNet: A Deep Learning Approach for Image Compressive Sensing (2020, methodology): “Recently, deep generative adversarial networks have been proposed for CS medical imaging in [51], [52], [53].”
CT Super-Resolution GAN Constrained by the Identical, Residual, and Cycle Learning Ensemble (GAN-CIRCLE) (2018, methodology): “fully developed [31]–[38].”
Robust Compressed Sensing MRI with Deep Generative Priors (2021, background): “The approach in [13] has been generalized to tackle different inverse problems [45, 34, 7, 67, 56, 59, 70, 9], and different reconstruction algorithms [21, 48, 65, 27, 26, 60, 36, 37, 18].”
Deep Learning for Multigrade Brain Tumor Classification in Smart Healthcare Systems: A Prospective Survey (2020, background): “…techniques and are widely used for many applications in diverse domains, such as medical image synthesis [127], [128], compressive sensing MRI [129], [130], superresolution [130], security [131], classi-ﬁcation [132], image-to-image transformation [133], and many other useful purposes.”
A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications (2020, background): “cal fields such as for generating and designing DNA [296], [297], drug discovery [298], generating multi-label discrete patient records [299], medical image processing [300], [301], [302], [303], [304], [305], [306], [307], and doctor recommen-”