Accelerating 4D image reconstruction for magnetic resonance-guided radiotherapy.

4D-MRI High-performance computing Intrafraction motion MR-guided Radiotherapy MR-integrated Proton Therapy

Journal

Physics and imaging in radiation oncology
ISSN: 2405-6316
Titre abrégé: Phys Imaging Radiat Oncol
Pays: Netherlands
ID NLM: 101704276

Informations de publication

Date de publication:
Jul 2023
Historique:
received: 13 02 2023
revised: 15 08 2023
accepted: 16 08 2023
medline: 4 9 2023
pubmed: 4 9 2023
entrez: 4 9 2023
Statut: epublish

Résumé

Physiological motion impacts the dose delivered to tumours and vital organs in external beam radiotherapy and particularly in particle therapy. The excellent soft-tissue demarcation of 4D magnetic resonance imaging (4D-MRI) could inform on intra-fractional motion, but long image reconstruction times hinder its use in online treatment adaptation. Here we employ techniques from high-performance computing to reduce 4D-MRI reconstruction times below two minutes to facilitate their use in MR-guided radiotherapy. Four patients with pancreatic adenocarcinoma were scanned with a radial stack-of-stars gradient echo sequence on a 1.5T MR-Linac. Fast parallelised open-source implementations of the extra-dimensional golden-angle radial sparse parallel algorithm were developed for central processing unit (CPU) and graphics processing unit (GPU) architectures. We assessed the impact of architecture, oversampling and respiratory binning strategy on 4D-MRI reconstruction time and compared images using the structural similarity (SSIM) index against a MATLAB reference implementation. Scaling and bottlenecks for the different architectures were studied using multi-GPU systems. All reconstructed 4D-MRI were identical to the reference implementation (SSIM Respiratory-resolved 4D-MRI reconstruction times can be reduced using high-performance computing methods for online workflows in MR-guided radiotherapy with potential applications in particle therapy.

Sections du résumé

Background and purpose UNASSIGNED
Physiological motion impacts the dose delivered to tumours and vital organs in external beam radiotherapy and particularly in particle therapy. The excellent soft-tissue demarcation of 4D magnetic resonance imaging (4D-MRI) could inform on intra-fractional motion, but long image reconstruction times hinder its use in online treatment adaptation. Here we employ techniques from high-performance computing to reduce 4D-MRI reconstruction times below two minutes to facilitate their use in MR-guided radiotherapy.
Material and methods UNASSIGNED
Four patients with pancreatic adenocarcinoma were scanned with a radial stack-of-stars gradient echo sequence on a 1.5T MR-Linac. Fast parallelised open-source implementations of the extra-dimensional golden-angle radial sparse parallel algorithm were developed for central processing unit (CPU) and graphics processing unit (GPU) architectures. We assessed the impact of architecture, oversampling and respiratory binning strategy on 4D-MRI reconstruction time and compared images using the structural similarity (SSIM) index against a MATLAB reference implementation. Scaling and bottlenecks for the different architectures were studied using multi-GPU systems.
Results UNASSIGNED
All reconstructed 4D-MRI were identical to the reference implementation (SSIM
Conclusion UNASSIGNED
Respiratory-resolved 4D-MRI reconstruction times can be reduced using high-performance computing methods for online workflows in MR-guided radiotherapy with potential applications in particle therapy.

Identifiants

DOI: 10.1016/j.phro.2023.100484 PMID: 37664799 PMC: PMC10474606
pubmed: 37664799
doi: 10.1016/j.phro.2023.100484
pii: S2405-6316(23)00075-1
pmc: PMC10474606
doi:

Types de publication

Journal Article

Langues

eng

Pagination

100484

Informations de copyright

© 2023 The Author(s).

Déclaration de conflit d'intérêts

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: The Institute of Cancer Research and the Royal Marsden NHS Foundation Trust are part of the Elekta MR-Linac Research Consortium. We thank Philips for partnering with us on this research and providing MR source code, research licences, and support.

Références

Phys Imaging Radiat Oncol. 2021 Jan 07;17:32-35
pubmed: 33898775
Phys Med Biol. 2018 Oct 23;63(21):21TR01
pubmed: 30272573
Med Phys. 2009 Jun;36(6):2084-8
pubmed: 19610297
Magn Reson Med. 2015 Jan;73(1):292-8
pubmed: 24478142
IEEE Trans Med Imaging. 2005 Jun;24(6):799-808
pubmed: 15959939
Magn Reson Med. 2022 Dec;88(6):2592-2608
pubmed: 36128894
Radiother Oncol. 2020 Apr;145:88-94
pubmed: 31931291
Phys Med Biol. 2015 Nov 21;60(22):R323-61
pubmed: 26509844
Magn Reson Med. 2016 Sep;76(3):757-68
pubmed: 26362967
Phys Med Biol. 2013 Jun 21;58(12):4137-56
pubmed: 23689035
Magn Reson Med. 2017 Mar;77(3):1170-1183
pubmed: 26991911
Radiat Oncol. 2019 Feb 7;14(1):30
pubmed: 30732657
J Magn Reson Imaging. 2014 Nov;40(5):1022-40
pubmed: 24408499
Magn Reson Med. 2020 Sep;84(3):1280-1292
pubmed: 32086858
Med Phys. 2017 Aug;44(8):e77-e90
pubmed: 28547820
Clin Transl Radiat Oncol. 2020 May 15;23:72-79
pubmed: 32490218
Clin Oncol (R Coll Radiol). 2018 Nov;30(11):711-719
pubmed: 30201276
Med Phys. 2023 Aug 1;:
pubmed: 37527331
Phys Med Biol. 2017 Apr 21;62(8):2910-2921
pubmed: 27997382
IEEE Trans Biomed Eng. 2009 Apr;56(4):1134-42
pubmed: 19174334
Magn Reson Med. 2007 Dec;58(6):1182-95
pubmed: 17969013
Semin Radiat Oncol. 2014 Jul;24(3):203-6
pubmed: 24931094
Magn Reson Med. 2000 May;43(5):682-90
pubmed: 10800033
Med Phys. 2007 Dec;34(12):4768-71
pubmed: 18196804
Eur Heart J Suppl. 2023 Apr 26;25(Suppl C):C265-C270
pubmed: 37125279
Phys Imaging Radiat Oncol. 2023 Jan 02;25:100409
pubmed: 36655213
AJR Am J Roentgenol. 2019 Jun;212(6):1361-1369
pubmed: 30860902
Radiother Oncol. 2022 May;170:37-47
pubmed: 35257848
Radiother Oncol. 2021 Jun;159:209-217
pubmed: 33812914
Radiat Oncol. 2020 May 29;15(1):129
pubmed: 32471500
Magn Reson Med. 2016 Feb;75(2):775-88
pubmed: 25809847
Phys Imaging Radiat Oncol. 2019 Sep 23;11:82-87
pubmed: 33458284
IEEE Trans Image Process. 2004 Apr;13(4):600-12
pubmed: 15376593
Phys Med Biol. 2022 Jun 27;67(13):
pubmed: 35697010
Phys Imaging Radiat Oncol. 2022 May 14;22:111-114
pubmed: 35619641
Magn Reson Med. 2010 Aug;64(2):457-71
pubmed: 20665790
Radiother Oncol. 2008 Jan;86(1):25-9
pubmed: 18023488
Phys Med Biol. 2016 Jul 21;61(14):5335-55
pubmed: 27362636
Magn Reson Med. 2003 Dec;50(6):1276-85
pubmed: 14648576

Auteurs

Bastien Lecoeur (B)

Joint Department of Physics at The Institute of Cancer Research and The Royal Marsden NHS Foundation Trust, 15 Cotswold Rd, London SM2 5NG, United Kingdom.
Department of Computing, Imperial College London, Exhibition Rd, South Kensington, London SW7 2BX, United Kingdom.

Marco Barbone (M)

Joint Department of Physics at The Institute of Cancer Research and The Royal Marsden NHS Foundation Trust, 15 Cotswold Rd, London SM2 5NG, United Kingdom.
Department of Computing, Imperial College London, Exhibition Rd, South Kensington, London SW7 2BX, United Kingdom.

Jessica Gough (J)

Department of Radiotherapy at the Royal Marsden NHS Foundation Trust, Downs Rd, London SM2 5PT, United Kingdom.

Uwe Oelfke (U)

Joint Department of Physics at The Institute of Cancer Research and The Royal Marsden NHS Foundation Trust, 15 Cotswold Rd, London SM2 5NG, United Kingdom.

Wayne Luk (W)

Department of Computing, Imperial College London, Exhibition Rd, South Kensington, London SW7 2BX, United Kingdom.

Georgi Gaydadjiev (G)

Department of Computing, Imperial College London, Exhibition Rd, South Kensington, London SW7 2BX, United Kingdom.
Bernoulli Institute, University of Groningen, Nijenborgh 9, Groningen 9747 AG, The Netherlands.

Andreas Wetscherek (A)

Joint Department of Physics at The Institute of Cancer Research and The Royal Marsden NHS Foundation Trust, 15 Cotswold Rd, London SM2 5NG, United Kingdom.

Classifications MeSH