Sampling possible reconstructions of undersampled acquisitions in MR imaging

  title={Sampling possible reconstructions of undersampled acquisitions in MR imaging},
  author={Kerem Can Tezcan and Christian F. Baumgartner and Ender Konukoglu},
  journal={IEEE transactions on medical imaging},
Undersampling the k-space during MR acquisitions saves time, however results in an ill-posed inversion problem, leading to an infinite set of images as possible solutions. Traditionally, this is tackled as a reconstruction problem by searching for a single "best" image out of this solution set according to some chosen regularization or prior. This approach, however, misses the possibility of other solutions and hence ignores the uncertainty in the inversion process. In this paper, we propose a… 
1 Citations
Bayesian Inversion for Nonlinear Imaging Models using Deep Generative Priors
Most modern imaging systems involve a computational reconstruction pipeline to infer the image of interest from acquired measurements. The Bayesian reconstruction framework relies on the


Joint reconstruction and bias field correction for undersampled MR imaging
This work addresses the sensitivity of the reconstruction problem to the bias field and proposes to model it explicitly in the reconstruction, in order to decrease this sensitivity, and shows that the proposed method improves the reconstruction quality, both visually and in terms of RMSE.
MR Image Reconstruction Using Deep Density Priors
This letter proposes an alternative approach that learns the probability distribution of fully sampled MR images using unsupervised DL, specifically variational autoencoders (VAE), and uses this as an explicit prior term in reconstruction, completely decoupling the encoding operation from the prior.
Bayesian Deep Learning for Accelerated MR Image Reconstruction
It is shown that the proposed Bayesian methods achieve competitive performance when the test images are relatively far from the training data distribution and outperforms when the baseline method is over-parametrised, demonstrating the potential utility of the Bayesian DL methods for assessing the reliability of the reconstructed images.
Inverse GANs for accelerated MRI reconstruction
An iterative reconstruction approach that incorporates the implicit prior provided by a generative adversarial network (GAN), which learns the probability distribution of uncorrupted MRI data in an off-line step, that is evaluated on multi-coil knee MRI data and compared to a classical as well as a deep learning-based approach.
A Deep Cascade of Convolutional Neural Networks for Dynamic MR Image Reconstruction
A framework for reconstructing dynamic sequences of 2-D cardiac magnetic resonance images from undersampled data using a deep cascade of convolutional neural networks (CNNs) to accelerate the data acquisition process is proposed and it is demonstrated that CNNs can learn spatio-temporal correlations efficiently by combining convolution and data sharing approaches.
Uncertainty Quantification in Deep Learning for Safer Neuroimage Enhancement
Methods to characterise different components of uncertainty in medical image enhancement problems and demonstrate the ideas using diffusion MRI super-resolution to highlight three key benefits of uncertainty modelling for improving the safety of DL-based image enhancement systems.
Deep ADMM-Net for Compressive Sensing MRI
Experiments on MRI image reconstruction under different sampling ratios in k-space demonstrate that the proposed novel ADMM-Net algorithm significantly improves the baseline ADMM algorithm and achieves high reconstruction accuracies with fast computational speed.
Sparse MRI: The application of compressed sensing for rapid MR imaging
Practical incoherent undersampling schemes are developed and analyzed by means of their aliasing interference and demonstrate improved spatial resolution and accelerated acquisition for multislice fast spin‐echo brain imaging and 3D contrast enhanced angiography.
Modelling the Distribution of 3D Brain MRI using a 2D Slice VAE
This work proposes a method to model 3D MR brain volumes distribution by combining a 2D slice VAE with a Gaussian model that captures the relationships between slices and introduces a novel evaluation method for generated volumes that quantifies how well their segmentations match those of true brain anatomy.
Advances in sensitivity encoding with arbitrary k‐space trajectories
Using the proposed method, SENSE becomes practical with nonstandard k‐space trajectories, enabling considerable scan time reduction with respect to mere gradient encoding, and the in vivo feasibility of non‐Cartesian SENSE imaging with iterative reconstruction is demonstrated.