Abstract
Object
To enable high-quality physics-guided deep learning (PG-DL) reconstruction of large-scale 3D non-Cartesian coronary MRI by overcoming challenges of hardware limitations and limited training data availability.
Materials and methods
While PG-DL has emerged as a powerful image reconstruction method, its application to large-scale 3D non-Cartesian MRI is hindered by hardware limitations and limited availability of training data. We combine several recent advances in deep learning and MRI reconstruction to tackle the former challenge, and we further propose a 2.5D reconstruction using 2D convolutional neural networks, which treat 3D volumes as batches of 2D images to train the network with a limited amount of training data. Both 3D and 2.5D variants of the PG-DL networks were compared to conventional methods for high-resolution 3D kooshball coronary MRI.
Results
Proposed PG-DL reconstructions of 3D non-Cartesian coronary MRI with 3D and 2.5D processing outperformed all conventional methods both quantitatively and qualitatively in terms of image assessment by an experienced cardiologist. The 2.5D variant further improved vessel sharpness compared to 3D processing, and scored higher in terms of qualitative image quality.
Discussion
PG-DL reconstruction of large-scale 3D non-Cartesian MRI without compromising image size or network complexity is achieved, and the proposed 2.5D processing enables high-quality reconstruction with limited training data.
Similar content being viewed by others
Data availability
In accordance with the institutional review board, the data acquired in this study contain person-sensitive information, and can only be shared in the context of scientific collaborations.
References
Thedens DR, Irarrazaval P, Sachs TS, Meyer CH, Nishimura DG (1999) Fast magnetic resonance coronary angiography with a three-dimensional stack of spirals trajectory. Magn Reson Med 41(6):1170–1179
Wright KL, Hamilton JI, Griswold MA, Gulani V, Seiberlich N (2014) Non-Cartesian parallel imaging reconstruction. J Magn Reson Imaging 40(5):1022–1040
Chen Y, Lo WC, Hamilton JI, Barkauskas K, Saybasili H, Wright KL et al (2018) Single breath-hold 3D cardiac T1 mapping using through-time spiral GRAPPA. NMR Biomed 31(6):e3923
Irarrazabal P, Nishimura DG (1995) Fast three dimensional magnetic resonance imaging. Magn Reson Med 33(5):656–662
Stehning C, Börnert P, Nehrke K, Eggers H, Dössel O (2004) Fast isotropic volumetric coronary MR angiography using free-breathing 3D radial balanced FFE acquisition. Magn Reson Med 52(1):197–203
Gurney PT, Hargreaves BA, Nishimura DG (2006) Design and analysis of a practical 3D cones trajectory. Magn Reson Med 55(3):575–582
Barger AV, Block WF, Toropov Y, Grist TM, Mistretta CA (2002) Time-resolved contrast-enhanced imaging with isotropic resolution and broad coverage using an undersampled 3D projection trajectory. Magn Reson Med 48(2):297–305
Feng L, Grimm R, Block KT, Chandarana H, Kim S, Xu J et al (2014) Golden-angle radial sparse parallel MRI: combination of compressed sensing, parallel imaging, and golden-angle radial sampling for fast and flexible dynamic volumetric MRI. Magn Reson Med 72(3):707–717
Piccini D, Feng L, Bonanno G, Coppo S, Yerly J, Lim RP et al (2017) Four-dimensional respiratory motion-resolved whole heart coronary MR angiography. Magn Reson Med 77(4):1473–1484
Bonanno G, Piccini D, Marchal B, Zenge M, Stuber M (2014) A new binning approach for 3D motion corrected self-navigated whole-heart coronary MRA using independent component analysis of individual coils. In: Proceedings of the 22nd annual meeting of ISMRM, Milan, p 936
Larson AC, White RD, Laub G, McVeigh ER, Li D, Simonetti OP (2004) Self-gated cardiac cine MRI. Magn Reson Med 51(1):93–102
Stehning C, Börnert P, Nehrke K, Dössel O (2005) Free breathing 3D balanced FFE coronary magnetic resonance angiography with prolonged cardiac acquisition windows and intra-RR motion correction. Magn Reson Med 53(3):719–723
Feng L, Coppo S, Piccini D, Yerly J, Lim RP, Masci PG et al (2018) 5D whole-heart sparse MRI. Magn Reson Med 79(2):826–838
Mistretta CA, Wieben O, Velikina J, Block W, Perry J, Wu Y et al (2006) Highly constrained backprojection for time-resolved MRI. Magn Reson Med 55(1):30–40
Feng L, Delacoste J, Smith D, Weissbrot J, Flagg E, Moore WH et al (2019) Simultaneous evaluation of lung anatomy and ventilation using 4D respiratory-motion-resolved ultrashort echo time sparse MRI. J Magn Reson Imaging 49(2):411–422
Nam S, Akcakaya M, Basha T, Stehning C, Manning WJ, Tarokh V et al (2013) Compressed sensing reconstruction for whole-heart imaging with 3D radial trajectories: a graphics processing unit implementation. Magn Reson Med 69(1):91–102
Hammernik K, Klatzer T, Kobler E, Recht MP, Sodickson DK, Pock T et al (2018) Learning a variational network for reconstruction of accelerated MRI data. Magn Reson Med 79(6):3055–3071
Schlemper J, Caballero J, Hajnal JV, Price A, Rueckert D (2017) A deep cascade of convolutional neural networks for MR image reconstruction. In: Information processing in medical imaging: 25th international conference, IPMI 2017, Boone, NC, USA, June 25–30, 2017, Proceedings 25. Springer, pp 647–658
Aggarwal HK, Mani MP, Jacob M (2018) MoDL: Model-based deep learning architecture for inverse problems. IEEE Trans Med Imaging 38(2):394–405
Knoll F, Hammernik K, Zhang C, Moeller S, Pock T, Sodickson DK et al (2020) Deep-learning methods for parallel magnetic resonance imaging reconstruction: a survey of the current approaches, trends, and issues. IEEE Signal Process Mag 37(1):128–140
Yaman B, Hosseini SAH, Moeller S, Ellermann J, Uğurbil K, Akçakaya M (2020) Self-supervised learning of physics-guided reconstruction neural networks without fully sampled reference data. Magn Reson Med 84(6):3172–3191
Hammernik K, Küstner T, Yaman B, Huang Z, Rueckert D, Knoll F et al (2023) Physics-driven deep learning for computational magnetic resonance imaging: combining physics and machine learning for improved medical imaging. IEEE Signal Process Mag 40(1):98–114
Piccini D, Littmann A, Nielles-Vallespin S, Zenge MO (2011) Spiral phyllotaxis: the natural way to construct a 3D radial trajectory in MRI. Magn Reson Med 66(4):1049–1056
Ramzi Z, Chaithya G, Starck J-L, Ciuciu P (2022) NC-PDNet: a density-compensated unrolled network for 2D and 3D non-Cartesian MRI reconstruction. IEEE Trans Med Imaging 41(7):1625–1638
Malavé MO, Baron CA, Koundinyan SP, Sandino CM, Ong F, Cheng JY et al (2020) Reconstruction of undersampled 3D non-Cartesian image-based navigators for coronary MRA using an unrolled deep learning model. Magn Reson Med 84(2):800–812
Deng Z, Yaman B, Zhang C, Moeller S, Akçakaya M (2021) Efficient training of 3D unrolled neural networks for MRI reconstruction using small databases. In: 2021 55th Asilomar conference on signals, systems, and computers. IEEE, pp 886–889.
Chen Z, Chen Y, Xie Y, Li D, Christodoulou AG (2022) Data-consistent non-Cartesian deep subspace learning for efficient dynamic MR image reconstruction. In: 2022 IEEE 19th international symposium on biomedical imaging (ISBI). IEEE, pp 1–5
Kellman M, Zhang K, Markley E, Tamir J, Bostan E, Lustig M et al (2020) Memory-efficient learning for large-scale computational imaging. IEEE Trans Comput Imaging 6:1403–1414
Baron CA, Dwork N, Pauly JM, Nishimura DG (2018) Rapid compressed sensing reconstruction of 3D non-Cartesian MRI. Magn Reson Med 79(5):2685–2692
Ramani S, Fessler JA (2013) Accelerated nonCartesian SENSE reconstruction using a majorize-minimize algorithm combining variable-splitting. In: 2013 IEEE 10th international symposium on biomedical imaging. IEEE, pp 704–707
Micikevicius P, Narang S, Alben J, Diamos G, Elsen E, Garcia D et al (2017) Mixed precision training. arXiv:1710.03740
Zhou L, Pan S, Wang J, Vasilakos AV (2017) Machine learning on big data: opportunities and challenges. Neurocomputing 237:350–361
Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P (1999) SENSE: sensitivity encoding for fast MRI. Magn Reson Med 42(5):952–962
Pruessmann KP, Weiger M, Börnert P, Boesiger P (2001) Advances in sensitivity encoding with arbitrary k-space trajectories. Magn Reson Med 46(4):638–651
Fessler JA (2020) Optimization methods for magnetic resonance image reconstruction: Key models and optimization algorithms. IEEE Signal Process Mag 37(1):33–40
Kumar R, Purohit M, Svitkina Z, Vee E, Wang J (2019) Efficient rematerialization for deep networks. In: Advances in neural information processing systems, p 32
Griewank A (1999) An implementation of checkpointing for the reverse or adjoint model of differentiation. ACM Trans Math Softw 26(1):1–19
Jain P, Jain A, Nrusimha A, Gholami A, Abbeel P, Gonzalez J et al (2020) Checkmate: breaking the memory wall with optimal tensor rematerialization. Proc Mach Learn Syst 2:497–511
Etienne A, Botnar RM, Van Muiswinkel AM, Boesiger P, Manning WJ, Stuber M (2002) “Soap-Bubble” visualization and quantitative analysis of 3D coronary magnetic resonance angiograms. Magn Reson Med 48(4):658–666
Hutchinson M, Raff U (1988) Fast MRI data acquisition using multiple detectors. Magn Reson Med 6(1):87–91
Ra JB, Rim C (1993) Fast imaging using subencoding data sets from multiple detectors. Magn Reson Med 30(1):142–145
Cruz G, Atkinson D, Henningsson M, Botnar RM, Prieto C (2017) Highly efficient nonrigid motion-corrected 3D whole-heart coronary vessel wall imaging. Magn Reson Med 77(5):1894–1908
Zhang C, Piccini D, Demirel OB, Bonanno G, Yaman B, Stuber M et al (2022) Distributed memory-efficient physics-guided deep learning reconstruction for large-scale 3d non-Cartesian MRI. In: 2022 IEEE 19th international symposium on biomedical imaging (ISBI). IEEE, pp 1–5
Muckley MJ, Stern R, Murrell T, Knoll F (2020) TorchKbNufft: a high-level, hardware-agnostic non-uniform fast Fourier transform. In: ISMRM workshop on data sampling & image reconstruction
Minnema J, Wolff J, Koivisto J, Lucka F, Batenburg KJ, Forouzanfar T et al (2021) Comparison of convolutional neural network training strategies for cone-beam CT image segmentation. Comput Methods Programs Biomed 207:106192
Bermejo-Peláez D, Estepar RSJ, Ledesma-Carbayo MJ (2018) Emphysema classification using a multi-view convolutional network. In: 2018 IEEE 15th international symposium on biomedical imaging (ISBI 2018). IEEE, pp 519–522
Ziabari A, Ye DH, Srivastava S, Sauer KD, Thibault J-B, Bouman CA (2018) 2.5 D deep learning for CT image reconstruction using a multi-GPU implementation. In: 2018 52nd Asilomar conference on signals, systems, and computers. IEEE, pp 2044–2049
Prasoon A, Petersen K, Igel C, Lauze F, Dam E, Nielsen M (2013) Deep feature learning for knee cartilage segmentation using a triplanar convolutional neural network. In: Medical image computing and computer-assisted intervention–MICCAI 2013: 16th international conference, Nagoya, Japan, September 22–26, 2013, Proceedings, Part II 16. Springer, pp 246–253
Shorten C, Khoshgoftaar TM (2019) A survey on image data augmentation for deep learning. J Big Data 6(1):1–48
Henningsson M, Koken P, Stehning C, Razavi R, Prieto C, Botnar RM (2012) Whole-heart coronary MR angiography with 2D self-navigated image reconstruction. Magn Reson Med 67(2):437–445
Küstner T, Munoz C, Psenicny A, Bustin A, Fuin N, Qi H et al (2021) Deep-learning based super-resolution for 3D isotropic coronary MR angiography in less than a minute. Magn Reson Med 86(5):2837–2852
Ahmad R, Ding Y, Simonetti OP (2015) Edge sharpness assessment by parametric modeling: application to magnetic resonance imaging. Concepts Magn Reson Part A 44(3):138–149
Gilton D, Ongie G, Willett R (2021) Deep equilibrium architectures for inverse problems in imaging. IEEE Trans Comput Imaging 7:1123–1133
Güngör A, Askin B, Soydan DA, Top CB, Saritas EU, Çukur T (2023) DEQ-MPI: a deep equilibrium reconstruction with learned consistency for magnetic particle imaging. IEEE Trans Med Imaging
Seiberlich N, Breuer F, Blaimer M, Jakob P, Griswold M (2008) Self-calibrating GRAPPA operator gridding for radial and spiral trajectories. Magn Reson Med 59(4):930–935
Akçakaya M, Nam S, Basha TA, Kawaji K, Tarokh V, Nezafat R (2014) An augmented Lagrangian based compressed sensing reconstruction for non-Cartesian magnetic resonance imaging without gridding and regridding at every iteration. PLoS ONE 9(9):e107107
Yurt M, Özbey M, Dar SU, Tinaz B, Oguz KK, Çukur T (2022) Progressively volumetrized deep generative models for data-efficient contextual learning of MR image recovery. Med Image Anal 78:102429
Dar SUH, Özbey M, Çatlı AB, Çukur T (2020) A transfer-learning approach for accelerated MRI using deep neural networks. Magn Reson Med 84(2):663–685
Funding
NIH, Grant numbers: R01HL153146, R01EB032830, R21EB028369, P41EB027061.
Author information
Authors and Affiliations
Contributions
Mehmet Akçakaya: Study conception and design, drafting of manuscript, critical revision. Gabriele Bonanno: Acquisition of data, critical revision. Omer Burak Demirel: Analysis and interpretation of data, critical revision. Steen Moeller: Analysis and interpretation of data, critical revision. Davide Piccini: Acquisition of data, critical revision. Burhaneddin Yaman: Analysis and interpretation of data, critical revision. Chetan Shenoy: Analysis and interpretation of data, critical revision. Matthias Stuber: Acquisition of data, critical revision. Chi Zhang: Study conception and design, drafting of manuscript, critical revision. Christopher W. Roy: Acquisition of data, critical revision.
Corresponding author
Ethics declarations
Conflict of interest
Davide Piccini and Gabriele Bonanno are employed by Siemens Healthineers AG, Bern, Switzerland. Matthias Stuber receives non-monetary research support from Siemens Healthineers that is covered by a master research agreement handled by his employer (CHUV). Matthias Stuber has a research contract with Circle that is handled by the Tech Transfer Office (PACTT) of his employer (CHUV).
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Zhang, C., Piccini, D., Demirel, O.B. et al. Large-scale 3D non-Cartesian coronary MRI reconstruction using distributed memory-efficient physics-guided deep learning with limited training data. Magn Reson Mater Phy (2024). https://doi.org/10.1007/s10334-024-01157-8
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s10334-024-01157-8