Learning a variational network for reconstruction of accelerated MRI data

Abstract

Variational network reconstructions preserve the natural appearance of MR images as well as pathologies that were not included in the training data set. Due to its high computational performance, that is, reconstruction time of 193 ms on a single graphics card, and the omission of parameter tuning once the network is trained, this new approach to image reconstruction can easily be integrated into clinical workflow. Magn Reson Med 79:3055-3071, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Real‐time cardiovascular MR with spatio‐temporal artifact suppression using deep learning–proof of concept in congenital heart disease

“…Assessment of image quality from the trained networks using the synthetic test data, described above, demonstrated <0.01% difference in RMSE and <0.5% difference in SSIM (further information in Supporting Information Figure S8). This is similar to previous deep learning studies, which have shown limited differences in RMSE and SSIM values using ‐loss or SSIM‐loss functions, however, with improved visual quality compared to ‐loss . Nevertheless, further investigation of different loss functions and other optimizations of learning would be desirable in the future.…”

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

“…This is significantly different from how human radiologists learn to read and interpret MRI images. Radiologists have been trained by reading thousands of MRI images to develop remarkable skills to recognise certain reproducible anatomical and contextual patterns in the images even with known artefacts presented [1], [44]. Our deep learning based DAGAN method aims to imitate this human learning procedure, and therefore shifts the conventional online nonlinear optimisation into an offline training procedure.…”

“…For the former learning scheme, there is lack of adaptivity, and for the latter one, the resulting dictionary in sparse coding is not hierarchical as in the deep learning based methods, which in general could provide superior results. In addition, the performance of our DAGAN method is also improved by enriching the training datasets with a comprehensive data augmentation that has not been considered in previous dictionary learning or deep learning based methods [26], [42], [36], [43], [44], [45]. Once a DAGAN model has been trained, it can be used to infer any new input raw data with the same undersampling ratio.…”

“…The method achieved similar reconstruction results as classic CS-MRI reconstruction methods but dramatically reduced reconstruction time. There are also three preprints describing deep learning based CS-MRI: Schlemper et al [43] proposed cascaded CNN incorporating a data consistency layer, Hammernik et al [44] trained a variational network to solve CS-MRI and Lee et al [45] combined CNN based CS-MRI with parallel imaging to estimate and remove the aliasing artefacts. Although deep learning has shown great potential in solving CS-MRI with much faster reconstruction, to date improvement was not found significantly different from what classic CS-MRI can achieve.…”

Single‐shot T2 mapping using overlapping‐echo detachment planar imaging and a deep convolutional neural network

“…Therefore, the mapping relationship can be learned by DNN if sufficient qualified training data can be obtained. However, compared with convex L 1 minimization where the characteristics can be understood more easily, deep‐learning based methods are often considered to be ‘black boxes’ and are still difficult to interpret …”