articleTop 1% cited

Learning a variational network for reconstruction of accelerated MRI data

Magnetic Resonance in Medicine · 2017 · Vol. 79(6) · pp. 3055–3071

Kerstin Hammernik✉(Graz University of Technology)Teresa Klatzer(Graz University of Technology)Erich Kobler(Graz University of Technology)Michael P. Recht(New York University)Daniel K. Sodickson(New York University)Thomas Pock(Austrian Institute of Technology)Florian Knöll(Advanced Imaging Research (United States))

View via DOI Publisher page

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.

Advanced MRI Techniques and Applications Medical Imaging Techniques and Applications Cardiac Imaging and DiagnosticsCompressed sensingComputer scienceAccelerationGradient descentArtificial intelligenceAlgorithmIterative reconstructionVariational methodDeep learningArtificial neural network

MeSH terms

Machine LearningAdolescentAdultAgedAlgorithmsComputer SimulationFemaleHumansImage Interpretation, Computer-AssistedImage Processing, Computer-AssistedMagnetic Resonance ImagingMaleMiddle AgedRetrospective StudiesReproducibility of Results