End-to-end deep learning nonrigid motion-corrected reconstruction for highly accelerated free-breathing coronary MRA

TY - JOUR

T1 - End-to-end deep learning nonrigid motion-corrected reconstruction for highly accelerated free-breathing coronary MRA

AU - Qi, Haikun

AU - Hajhosseiny, Reza

AU - Cruz, Gastao

AU - Kuestner, Thomas

AU - Kunze, Karl

AU - Neji, Radhouene

AU - Botnar, René

AU - Prieto, Claudia

N1 - Funding Information: Supported by The Engineering and Physical Sciences Research Council (EPSRC), grants: EP/P032311/1, EP/P001009/1, and EP/P007619/1; British Heart Foundation (BHF) grant PG/18/59/33955 and RG/20/1/34802; King’s BHF Centre for Research Excellence RE/18/2/34213; Wellcome EPSRC Centre for Medical Engineering grant NS/A000049/1; and the Department of Health via the National Institute for Health Research (NIHR) Cardiovascular Health Technology Cooperative (HTC) and comprehensive Biomedical Research Centre awarded to Guy’s & St Thomas’ NHS Foundation Trust in partnership with King’s College London and King’s College Hospital NHS Foundation Trust Funding Information: This work was supported by the following grants: (1) EPSRC EP/P032311/1, EP/P001009/1, and EP/P007619/1; (2) BHF PG/18/59/33955 and RG/20/1/34802; (3) King’s BHF Centre for Research Excellence RE/18/2/34213; (4) Wellcome EPSRC Centre for Medical Engineering (NS/A000049/1); and (5) the Department of Health via the National Institute for Health Research (NIHR) Cardiovascular Health Technology Cooperative (HTC) and comprehensive Biomedical Research Centre awarded to Guy’s & St Thomas’ NHS Foundation Trust in partnership with King’s College London and King’s College Hospital NHS Foundation Trust. Publisher Copyright: © 2021 International Society for Magnetic Resonance in Medicine Copyright: Copyright 2021 Elsevier B.V., All rights reserved.

PY - 2021/10

Y1 - 2021/10

N2 - Purpose: To develop an end-to-end deep learning technique for nonrigid motion-corrected (MoCo) reconstruction of ninefold undersampled free-breathing whole-heart coronary MRA (CMRA). Methods: A novel deep learning framework was developed consisting of a diffeomorphic registration network and a motion-informed model-based deep learning (MoDL) reconstruction network. The registration network receives as input highly undersampled (~22×) respiratory-resolved images and outputs 3D nonrigid respiratory motion fields between the images. The motion-informed MoDL performs MoCo reconstruction from undersampled data using the predicted motion fields. The whole deep learning framework, termed as MoCo-MoDL, was trained end-to-end in a supervised manner for simultaneous 3D nonrigid motion estimation and MoCo reconstruction. MoCo-MoDL was compared with a state-of-the-art nonrigid MoCo CMRA reconstruction technique in 15 retrospectively undersampled datasets and 9 prospectively undersampled acquisitions. Results: The acquisition time for ninefold accelerated CMRA was ~2.5 min. The reconstruction time was ~22 s for the proposed MoCo-MoDL and ~35 min for the conventional approach. MoCo-MoDL achieved higher peak SNR (27.86 ± 3.00 vs. 26.71 ± 2.79; P <.05) and structural similarity (0.78 ± 0.06 vs. 0.75 ± 0.06; P <.05) than the conventional approach. Similar vessel length and visual image quality score were obtained with the 2 methods, whereas improved vessel sharpness was observed with MoCo-MoDL. Conclusion: An end-to-end deep learning approach was introduced for simultaneous nonrigid motion estimation and MoCo reconstruction of highly undersampled free-breathing whole-heart CMRA. The rapid free-breathing CMRA acquisition together with the fast reconstruction of the proposed approach promises easy integration into clinical workflow.

AB - Purpose: To develop an end-to-end deep learning technique for nonrigid motion-corrected (MoCo) reconstruction of ninefold undersampled free-breathing whole-heart coronary MRA (CMRA). Methods: A novel deep learning framework was developed consisting of a diffeomorphic registration network and a motion-informed model-based deep learning (MoDL) reconstruction network. The registration network receives as input highly undersampled (~22×) respiratory-resolved images and outputs 3D nonrigid respiratory motion fields between the images. The motion-informed MoDL performs MoCo reconstruction from undersampled data using the predicted motion fields. The whole deep learning framework, termed as MoCo-MoDL, was trained end-to-end in a supervised manner for simultaneous 3D nonrigid motion estimation and MoCo reconstruction. MoCo-MoDL was compared with a state-of-the-art nonrigid MoCo CMRA reconstruction technique in 15 retrospectively undersampled datasets and 9 prospectively undersampled acquisitions. Results: The acquisition time for ninefold accelerated CMRA was ~2.5 min. The reconstruction time was ~22 s for the proposed MoCo-MoDL and ~35 min for the conventional approach. MoCo-MoDL achieved higher peak SNR (27.86 ± 3.00 vs. 26.71 ± 2.79; P <.05) and structural similarity (0.78 ± 0.06 vs. 0.75 ± 0.06; P <.05) than the conventional approach. Similar vessel length and visual image quality score were obtained with the 2 methods, whereas improved vessel sharpness was observed with MoCo-MoDL. Conclusion: An end-to-end deep learning approach was introduced for simultaneous nonrigid motion estimation and MoCo reconstruction of highly undersampled free-breathing whole-heart CMRA. The rapid free-breathing CMRA acquisition together with the fast reconstruction of the proposed approach promises easy integration into clinical workflow.

KW - coronary MRA

KW - deep learning nonrigid motion correction

KW - deep learning reconstruction

KW - free-breathing cardiac MRI

UR - http://www.scopus.com/inward/record.url?scp=85107230310&partnerID=8YFLogxK

U2 - 10.1002/mrm.28851

DO - 10.1002/mrm.28851

M3 - Article

AN - SCOPUS:85107230310

SN - 0740-3194

VL - 86

SP - 1983

EP - 1996

JO - Magnetic Resonance in Medicine

JF - Magnetic Resonance in Medicine

IS - 4

ER -