VORTEX: Physics-Driven Data Augmentations Using Consistency Training for Robust Accelerated MRI Reconstruction

Arjun D. Desai, Beliz Gunel, Batu M. Ozturkler, Harris Beg, Shreyas Vasanawala, Brian A. Hargreaves, Christopher Ré, John M Pauly, Akshay S. Chaudhari
Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, PMLR 172:325-352, 2022.

Abstract

Deep neural networks have enabled improved image quality and fast inference times for various inverse problems, including accelerated magnetic resonance imaging (MRI) reconstruction. However, such models require a large number of fully-sampled ground truth datasets, which are difficult to curate, and are sensitive to distribution drifts. In this work, we propose applying physics-driven data augmentations for consistency training that leverage our domain knowledge of the forward MRI data acquisition process and MRI physics to achieve improved label efficiency and robustness to clinically-relevant distribution drifts. Our approach, termed VORTEX, (1) demonstrates strong improvements over supervised baselines with and without data augmentation in robustness to signal-to-noise ratio change and motion corruption in data-limited regimes; (2) considerably outperforms state-of-the-art purely image-based data augmentation techniques and self-supervised reconstruction methods on both in-distribution and out-of-distribution data; and (3) enables composing heterogeneous image-based and physics-driven data augmentations.

Cite this Paper

BibTeX
@InProceedings{pmlr-v172-desai22a, title = {VORTEX: Physics-Driven Data Augmentations Using Consistency Training for Robust Accelerated MRI Reconstruction}, author = {Desai, Arjun D. and Gunel, Beliz and Ozturkler, Batu M. and Beg, Harris and Vasanawala, Shreyas and Hargreaves, Brian A. and R{\'e}, Christopher and Pauly, John M and Chaudhari, Akshay S.}, booktitle = {Proceedings of The 5th International Conference on Medical Imaging with Deep Learning}, pages = {325--352}, year = {2022}, editor = {Konukoglu, Ender and Menze, Bjoern and Venkataraman, Archana and Baumgartner, Christian and Dou, Qi and Albarqouni, Shadi}, volume = {172}, series = {Proceedings of Machine Learning Research}, month = {06--08 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v172/desai22a/desai22a.pdf}, url = {https://proceedings.mlr.press/v172/desai22a.html}, abstract = {Deep neural networks have enabled improved image quality and fast inference times for various inverse problems, including accelerated magnetic resonance imaging (MRI) reconstruction. However, such models require a large number of fully-sampled ground truth datasets, which are difficult to curate, and are sensitive to distribution drifts. In this work, we propose applying physics-driven data augmentations for consistency training that leverage our domain knowledge of the forward MRI data acquisition process and MRI physics to achieve improved label efficiency and robustness to clinically-relevant distribution drifts. Our approach, termed VORTEX, (1) demonstrates strong improvements over supervised baselines with and without data augmentation in robustness to signal-to-noise ratio change and motion corruption in data-limited regimes; (2) considerably outperforms state-of-the-art purely image-based data augmentation techniques and self-supervised reconstruction methods on both in-distribution and out-of-distribution data; and (3) enables composing heterogeneous image-based and physics-driven data augmentations.} }
Endnote
%0 Conference Paper %T VORTEX: Physics-Driven Data Augmentations Using Consistency Training for Robust Accelerated MRI Reconstruction %A Arjun D. Desai %A Beliz Gunel %A Batu M. Ozturkler %A Harris Beg %A Shreyas Vasanawala %A Brian A. Hargreaves %A Christopher Ré %A John M Pauly %A Akshay S. Chaudhari %B Proceedings of The 5th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2022 %E Ender Konukoglu %E Bjoern Menze %E Archana Venkataraman %E Christian Baumgartner %E Qi Dou %E Shadi Albarqouni %F pmlr-v172-desai22a %I PMLR %P 325--352 %U https://proceedings.mlr.press/v172/desai22a.html %V 172 %X Deep neural networks have enabled improved image quality and fast inference times for various inverse problems, including accelerated magnetic resonance imaging (MRI) reconstruction. However, such models require a large number of fully-sampled ground truth datasets, which are difficult to curate, and are sensitive to distribution drifts. In this work, we propose applying physics-driven data augmentations for consistency training that leverage our domain knowledge of the forward MRI data acquisition process and MRI physics to achieve improved label efficiency and robustness to clinically-relevant distribution drifts. Our approach, termed VORTEX, (1) demonstrates strong improvements over supervised baselines with and without data augmentation in robustness to signal-to-noise ratio change and motion corruption in data-limited regimes; (2) considerably outperforms state-of-the-art purely image-based data augmentation techniques and self-supervised reconstruction methods on both in-distribution and out-of-distribution data; and (3) enables composing heterogeneous image-based and physics-driven data augmentations.
APA
Desai, A.D., Gunel, B., Ozturkler, B.M., Beg, H., Vasanawala, S., Hargreaves, B.A., Ré, C., Pauly, J.M. & Chaudhari, A.S.. (2022). VORTEX: Physics-Driven Data Augmentations Using Consistency Training for Robust Accelerated MRI Reconstruction. Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 172:325-352 Available from https://proceedings.mlr.press/v172/desai22a.html.

Related Material