Homotopic Gradients of Generative Density Priors for MR Image Reconstruction (2020-08-14T00:00:00.000000Z)

TL;DR

This work trains a more powerful noise conditional score network by forming high-dimensional tensor as the network input at the training phase and estimates the target gradients in higher-dimensional space to tackle low-dimensional manifold and low data density region issues in generative density prior.

Abstract

Deep learning, particularly the generative model, has demonstrated tremendous potential to significantly speed up image reconstruction with reduced measurements recently. Rather than the existing generative models that often optimize the density priors, in this work, by taking advantage of the denoising score matching, homotopic gradients of generative density priors (HGGDP) are exploited for magnetic resonance imaging (MRI) reconstruction. More precisely, to tackle the low-dimensional manifold and low data density region issues in generative density prior, we estimate the target gradients in higher-dimensional space. We train a more powerful noise conditional score network by forming high-dimensional tensor as the network input at the training phase. More artificial noise is also injected in the embedding space. At the reconstruction stage, a homotopy method is employed to pursue the density prior, such as to boost the reconstruction performance. Experiment results implied the remarkable performance of HGGDP in terms of high reconstruction accuracy. Only 10% of the k-space data can still generate image of high quality as effectively as standard MRI reconstructions with the fully sampled data.

Authors

Shanshan Wang

2 papers

D. Liang

2 papers

Cong Quan

1 papers

TL;DR

Abstract

Authors

References55 items

High-dimensional embedding network derived prior for compressive sensing MRI reconstruction

Generative Modeling with Denoising Auto-Encoders and Langevin Sampling

Annealed Denoising Score Matching: Learning Energy-Based Models in High-Dimensional Spaces

IFR-Net: Iterative Feature Refinement Network for Compressed Sensing MRI

MRI reconstruction using deep Bayesian estimation

Highly undersampled magnetic resonance imaging reconstruction using autoencoding priors

Generative Modeling by Estimating Gradients of the Data Distribution

Neural Empirical Bayes

SEGAN: Structure-Enhanced Generative Adversarial Network for Compressed Sensing MRI Reconstruction

Glow: Generative Flow with Invertible 1x1 Convolutions

DAGAN: Deep De-Aliasing Generative Adversarial Networks for Fast Compressed Sensing MRI Reconstruction

KIKI‐net: cross‐domain convolutional neural networks for reconstructing undersampled magnetic resonance images

Task-Aware Compressed Sensing with Generative Adversarial Networks

Lung and Pancreatic Tumor Characterization in the Deep Learning Era: Novel Supervised and Unsupervised Learning Approaches

MoDL: Model-Based Deep Learning Architecture for Inverse Problems

MR Image Reconstruction Using Deep Density Priors

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

Deep Mean-Shift Priors for Image Restoration

Autoencoder-based feature learning for cyber security applications

A Deep Cascade of Convolutional Neural Networks for Dynamic MR Image Reconstruction

Image Restoration using Autoencoding Priors

A Deep Cascade of Convolutional Neural Networks for MR Image Reconstruction

Feature Selection Guided Auto-Encoder

RefineNet: Multi-path Refinement Networks for High-Resolution Semantic Segmentation

Decoupled Algorithm for MRI Reconstruction Using Nonlocal Block Matching Model: BM3D-MRI

InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets

Accelerating magnetic resonance imaging via deep learning

Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks

U-Net: Convolutional Networks for Biomedical Image Segmentation

Magnetic resonance image reconstruction from undersampled measurements using a patch-based nonlocal operator

Compressive Sensing via Nonlocal Low-Rank Regularization

Auto-Encoding Variational Bayes

Adaptive Dictionary Learning in Sparse Gradient Domain for Image Recovery

Compressed Sensing-Based MRI Reconstruction Using Complex Double-Density Dual-Tree DWT

Highly Undersampled Magnetic Resonance Image Reconstruction Using Two-Level Bregman Method With Dictionary Updating

Sparse Reconstruction of Breast MRI Using Homotopic $L_0$ Minimization in a Regional Sparsified Domain

Bayesian Nonparametric Dictionary Learning for Compressed Sensing MRI

Undersampled MRI reconstruction with patch-based directional wavelets.

Representation Learning: A Review and New Perspectives

The Manifold Tangent Classifier

A Connection Between Score Matching and Denoising Autoencoders

MR Image Reconstruction From Highly Undersampled k-Space Data by Dictionary Learning

Second order total generalized variation (TGV) for MRI

Sample Complexity of Testing the Manifold Hypothesis

Three dimension double inversion recovery gray matter imaging using compressed sensing.

A Fast Alternating Direction Method for TVL1-L2 Signal Reconstruction From Partial Fourier Data

The Split Bregman Method for L1-Regularized Problems

Extracting and composing robust features with denoising autoencoders

Sparse MRI: The application of compressed sensing for rapid MR imaging

GENERATIVE ADVERSARIAL NETS

Image reconstruction of compressed sensing MRI using graph-based redundant wavelet transform

Efficient Algorithms for Convolutional Sparse Representations

Deep Learning

Compressed Sensing

Highly Undersampled Magnetic Resonance Image Reconstruction via Homotopic $\ell_{0}$ -Minimization

Field of Study

Journal Information

Name

Page

Volume

Venue Information

Name

Type

URL

Alternate Names