On learning adaptive acquisition policies for undersampled multi-coil MRI reconstruction

Tim Bakker, Matthew Muckley, Adriana Romero-Soriano, Michal Drozdzal, Luis Pineda
Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, PMLR 172:63-85, 2022.

Abstract

Most current approaches to undersampled multi-coil MRI reconstruction focus on learning the reconstruction model for a fixed, equidistant acquisition trajectory. In this paper, we study the problem of joint learning of the reconstruction model together with acquisition policies. To this end, we extend the End-to-End Variational Network with learnable acquisition policies that can adapt to different data points. We validate our model on a coil-compressed version of the large scale undersampled multi-coil fastMRI dataset using two undersampling factors: $4\times$ and $8\times$. Our experiments show on-par performance with the learnable non-adaptive and handcrafted equidistant strategies at $4\times$, and an observed improvement of more than $2%$ in SSIM at $8\times$ acceleration, suggesting that potentially-adaptive $k$-space acquisition trajectories can improve reconstructed image quality for larger acceleration factors. However, and perhaps surprisingly, our best performing policies learn to be explicitly non-adaptive.

Cite this Paper

BibTeX
@InProceedings{pmlr-v172-bakker22a, title = {On learning adaptive acquisition policies for undersampled multi-coil MRI reconstruction}, author = {Bakker, Tim and Muckley, Matthew and Romero-Soriano, Adriana and Drozdzal, Michal and Pineda, Luis}, booktitle = {Proceedings of The 5th International Conference on Medical Imaging with Deep Learning}, pages = {63--85}, 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/bakker22a/bakker22a.pdf}, url = {https://proceedings.mlr.press/v172/bakker22a.html}, abstract = {Most current approaches to undersampled multi-coil MRI reconstruction focus on learning the reconstruction model for a fixed, equidistant acquisition trajectory. In this paper, we study the problem of joint learning of the reconstruction model together with acquisition policies. To this end, we extend the End-to-End Variational Network with learnable acquisition policies that can adapt to different data points. We validate our model on a coil-compressed version of the large scale undersampled multi-coil fastMRI dataset using two undersampling factors: $4\times$ and $8\times$. Our experiments show on-par performance with the learnable non-adaptive and handcrafted equidistant strategies at $4\times$, and an observed improvement of more than $2%$ in SSIM at $8\times$ acceleration, suggesting that potentially-adaptive $k$-space acquisition trajectories can improve reconstructed image quality for larger acceleration factors. However, and perhaps surprisingly, our best performing policies learn to be explicitly non-adaptive.} }
Endnote
%0 Conference Paper %T On learning adaptive acquisition policies for undersampled multi-coil MRI reconstruction %A Tim Bakker %A Matthew Muckley %A Adriana Romero-Soriano %A Michal Drozdzal %A Luis Pineda %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-bakker22a %I PMLR %P 63--85 %U https://proceedings.mlr.press/v172/bakker22a.html %V 172 %X Most current approaches to undersampled multi-coil MRI reconstruction focus on learning the reconstruction model for a fixed, equidistant acquisition trajectory. In this paper, we study the problem of joint learning of the reconstruction model together with acquisition policies. To this end, we extend the End-to-End Variational Network with learnable acquisition policies that can adapt to different data points. We validate our model on a coil-compressed version of the large scale undersampled multi-coil fastMRI dataset using two undersampling factors: $4\times$ and $8\times$. Our experiments show on-par performance with the learnable non-adaptive and handcrafted equidistant strategies at $4\times$, and an observed improvement of more than $2%$ in SSIM at $8\times$ acceleration, suggesting that potentially-adaptive $k$-space acquisition trajectories can improve reconstructed image quality for larger acceleration factors. However, and perhaps surprisingly, our best performing policies learn to be explicitly non-adaptive.
APA
Bakker, T., Muckley, M., Romero-Soriano, A., Drozdzal, M. & Pineda, L.. (2022). On learning adaptive acquisition policies for undersampled multi-coil MRI reconstruction. Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 172:63-85 Available from https://proceedings.mlr.press/v172/bakker22a.html.

Related Material