Towards multi-sequence MR image recovery from undersampled k-space data

Cheng Peng, Wei-An Lin, Rama Chellappa, S. Kevin Zhou
; Proceedings of the Third Conference on Medical Imaging with Deep Learning, PMLR 121:614-623, 2020.

Abstract

Undersampled MR image recovery has been widely studied with Deep Learning methods as a post-processing step for accelerating MR acquisition. In this paper, we aim to optimize multi-sequence MR image recovery from undersampled k-space data under an overall time constraint. We first formulate it as a {\em constrained optimization} problem and show that finding the optimal sampling strategy for all sequences and the optimal recovery model for such sampling strategy is {\em combinatorial} and hence computationally prohibitive. To solve this problem, we propose a {\em blind recovery model} that simultaneously recovers multiple sequences, and an efficient approach to find proper combination of sampling strategy and recovery model. Our experiments demonstrate that the proposed method outperforms sequence-wise recovery, and sheds light on how to decide the undersampling strategy for sequences within an overall time budget.

Cite this Paper


BibTeX
@InProceedings{pmlr-v121-peng20a, title = {Towards multi-sequence MR image recovery from undersampled k-space data}, author = {Peng, Cheng and Lin, Wei-An and Chellappa, Rama and Zhou, S. Kevin}, pages = {614--623}, year = {2020}, editor = {Tal Arbel and Ismail Ben Ayed and Marleen de Bruijne and Maxime Descoteaux and Herve Lombaert and Christopher Pal}, volume = {121}, series = {Proceedings of Machine Learning Research}, address = {Montreal, QC, Canada}, month = {06--08 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v121/peng20a/peng20a.pdf}, url = {http://proceedings.mlr.press/v121/peng20a.html}, abstract = {Undersampled MR image recovery has been widely studied with Deep Learning methods as a post-processing step for accelerating MR acquisition. In this paper, we aim to optimize multi-sequence MR image recovery from undersampled k-space data under an overall time constraint. We first formulate it as a {\em constrained optimization} problem and show that finding the optimal sampling strategy for all sequences and the optimal recovery model for such sampling strategy is {\em combinatorial} and hence computationally prohibitive. To solve this problem, we propose a {\em blind recovery model} that simultaneously recovers multiple sequences, and an efficient approach to find proper combination of sampling strategy and recovery model. Our experiments demonstrate that the proposed method outperforms sequence-wise recovery, and sheds light on how to decide the undersampling strategy for sequences within an overall time budget.} }
Endnote
%0 Conference Paper %T Towards multi-sequence MR image recovery from undersampled k-space data %A Cheng Peng %A Wei-An Lin %A Rama Chellappa %A S. Kevin Zhou %B Proceedings of the Third Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2020 %E Tal Arbel %E Ismail Ben Ayed %E Marleen de Bruijne %E Maxime Descoteaux %E Herve Lombaert %E Christopher Pal %F pmlr-v121-peng20a %I PMLR %J Proceedings of Machine Learning Research %P 614--623 %U http://proceedings.mlr.press %V 121 %W PMLR %X Undersampled MR image recovery has been widely studied with Deep Learning methods as a post-processing step for accelerating MR acquisition. In this paper, we aim to optimize multi-sequence MR image recovery from undersampled k-space data under an overall time constraint. We first formulate it as a {\em constrained optimization} problem and show that finding the optimal sampling strategy for all sequences and the optimal recovery model for such sampling strategy is {\em combinatorial} and hence computationally prohibitive. To solve this problem, we propose a {\em blind recovery model} that simultaneously recovers multiple sequences, and an efficient approach to find proper combination of sampling strategy and recovery model. Our experiments demonstrate that the proposed method outperforms sequence-wise recovery, and sheds light on how to decide the undersampling strategy for sequences within an overall time budget.
APA
Peng, C., Lin, W., Chellappa, R. & Zhou, S.K.. (2020). Towards multi-sequence MR image recovery from undersampled k-space data. Proceedings of the Third Conference on Medical Imaging with Deep Learning, in PMLR 121:614-623

Related Material