Compressed Sensing MRI Reconstruction Using a Generative Adversarial Network With a Cyclic Loss

By Tran Minh Quan, Thanh Nguyen-Duc, Won-Ki Jeong. IEEE Transactions on Medical Imaging, 2017

588 citations recorded by the source corpus.

Compressed sensing magnetic resonance imaging (CS-MRI) has provided theoretical foundations upon which the time-consuming MRI acquisition process can be accelerated. However, it primarily relies on iterative numerical solvers, which still hinders their adaptation in time-critical applications. In addition, recent advances in deep neural networks have shown their potential in computer vision and image processing, but their adaptation to MRI reconstruction is still in an early stage. In this paper, we propose a novel deep learning-based generative adversarial model, <italic>RefineGAN</italic>, for fast and accurate CS-MRI reconstruction. The proposed model is a variant of fully-residual convolutional autoencoder and generative adversarial networks (GANs), specifically designed for CS-MRI formulation; it employs deeper generator and discriminator networks with cyclic data consistency loss for faithful interpolation in the given under-sampled <inline-formula> <tex-math notation="LaTeX">$k$ </tex-math></inline-formula>-space data. In addition, our solution leverages a chained network to further enhance the reconstruction quality. <italic>RefineGAN</italic> is fast and accurate—the reconstruction process is extremely rapid, as low as tens of milliseconds for reconstruction of a <inline-formula> <tex-math notation="LaTeX">$256\times 256$ </tex-math></inline-formula> image, because it is one-way deployment on a feed-forward network, and the image quality is superior even for extremely low sampling rate (as low as 10%) due to the data-driven nature of the method. We demonstrate that <italic>RefineGAN</italic> outperforms the state-of-the-art CS-MRI methods by a large margin in terms of both running time and image quality via evaluation using several open-source MRI databases.

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

View Paper: https://www.semanticscholar.org/paper/4652e86e535a15b191fcbe5c047e5ee0b9c86f09

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

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].”
MedGAN: Medical Image Translation using GANs (2018, methodology): “In order to circumvent this issue, current approaches utilize either specialized architectures for a given medical transformation task [51] or require the training of several frameworks one after the other [46].”
Distribution Matching Losses Can Hallucinate Features in Medical Image Translation (2018, methodology): “Applications include MR to CT [7, 8], CS-MRI [9, 10], CT to PET [11], and automatic H&E staining [12].”
Improving GANs for Speech Enhancement (2020, background): “[23] showed that using an additional generator chained to the generator of a GAN leads to better imagereconstruction performance.”
Deep Learning Techniques for Inverse Problems in Imaging (2020, background): “A similar idea was explored for MRI by [153], although in this case the forward operator is assumed to be a subsampled Fourier transform.”
Manipulating Medical Image Translation with Manifold Disentanglement (2020, background): “For example, segmentation map to medical image [20], [21], MR reconstruction [22], [23], image denoising [24], [25] and cross-modality translation [26], [27], [28].”
Noise2Noise: Learning Image Restoration without Clean Data (2018, background): “AmbientGAN (Ashish Bora, 2018) trains generative adversarial networks (Goodfellow et al., 2014) using corrupted observations.”
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-”
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.”
GANs for Medical Image Analysis (2018, methodology): “nd cancer because of the radiation exposure, which motivates the synthesis of CT images 27 MRI images GAN Patch extraction GAN É CT image CT image Patch extraction Figure 14: Proposed architecture in [22] from MR acquisitions. Nie et al. [76] synthesize CT images from corresponding MR images with th”
Solving inverse problems using data-driven models (2019, methodology): “This was done by Quinto (1993) for the case when the two- or three-dimensional ray transform is restricted to parallel lines, and by Quinto and Öktem (2008) for an analysis in the region-ofinterest limited angle setting.”