Compressed sensing dynamic MRI reconstruction using GPU-accelerated 3D convolutional sparse coding

Tran Minh Quan; Won Ki Jeong

doi:10.1007/978-3-319-46726-9_56

Compressed sensing dynamic MRI reconstruction using GPU-accelerated 3D convolutional sparse coding

Tran Minh Quan
, Won Ki Jeong^*

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

18 Citations (Scopus)

Abstract

In this paper,we introduce a fast alternating method for reconstructing highly undersampled dynamic MRI data using 3D convolutional sparse coding. The proposed solution leverages Fourier Convolution Theorem to accelerate the process of learning a set of 3D filters and iteratively refine the MRI reconstruction based on the sparse codes found subsequently. In contrast to conventional CS methods which exploit the sparsity by applying universal transforms such as wavelet and total variation,our approach extracts and adapts the temporal information directly from the MRI data using compact shift-invariant 3D filters. We provide a highly parallel algorithm with GPU support for efficient computation,and therefore,the reconstruction outperforms CPU implementation of the state-of-the art dictionary learning-based approaches by up to two orders of magnitude.

Original language	English
Title of host publication	Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings
Editors	Leo Joskowicz, Mert R. Sabuncu, William Wells, Gozde Unal, Sebastian Ourselin
Publisher	Springer Verlag
Pages	484-492
Number of pages	9
ISBN (Print)	9783319467252
DOIs	https://doi.org/10.1007/978-3-319-46726-9_56
Publication status	Published - 2016
Externally published	Yes

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9902 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Bibliographical note

Publisher Copyright:
© Springer International Publishing AG 2016.

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-319-46726-9_56

Cite this

Quan, T. M., & Jeong, W. K. (2016). Compressed sensing dynamic MRI reconstruction using GPU-accelerated 3D convolutional sparse coding. In L. Joskowicz, M. R. Sabuncu, W. Wells, G. Unal, & S. Ourselin (Eds.), Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings (pp. 484-492). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9902 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-46726-9_56

Compressed sensing dynamic MRI reconstruction using GPU-accelerated 3D convolutional sparse coding. / Quan, Tran Minh; Jeong, Won Ki.
Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings. ed. / Leo Joskowicz; Mert R. Sabuncu; William Wells; Gozde Unal; Sebastian Ourselin. Springer Verlag, 2016. p. 484-492 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9902 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Quan, TM & Jeong, WK 2016, Compressed sensing dynamic MRI reconstruction using GPU-accelerated 3D convolutional sparse coding. in L Joskowicz, MR Sabuncu, W Wells, G Unal & S Ourselin (eds), Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9902 LNCS, Springer Verlag, pp. 484-492. https://doi.org/10.1007/978-3-319-46726-9_56

Quan TM, Jeong WK. Compressed sensing dynamic MRI reconstruction using GPU-accelerated 3D convolutional sparse coding. In Joskowicz L, Sabuncu MR, Wells W, Unal G, Ourselin S, editors, Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings. Springer Verlag. 2016. p. 484-492. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-46726-9_56

Quan, Tran Minh ; Jeong, Won Ki. / Compressed sensing dynamic MRI reconstruction using GPU-accelerated 3D convolutional sparse coding. Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings. editor / Leo Joskowicz ; Mert R. Sabuncu ; William Wells ; Gozde Unal ; Sebastian Ourselin. Springer Verlag, 2016. pp. 484-492 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{f2e287886d23462598431d89a5ebcbae,

title = "Compressed sensing dynamic MRI reconstruction using GPU-accelerated 3D convolutional sparse coding",

abstract = "In this paper,we introduce a fast alternating method for reconstructing highly undersampled dynamic MRI data using 3D convolutional sparse coding. The proposed solution leverages Fourier Convolution Theorem to accelerate the process of learning a set of 3D filters and iteratively refine the MRI reconstruction based on the sparse codes found subsequently. In contrast to conventional CS methods which exploit the sparsity by applying universal transforms such as wavelet and total variation,our approach extracts and adapts the temporal information directly from the MRI data using compact shift-invariant 3D filters. We provide a highly parallel algorithm with GPU support for efficient computation,and therefore,the reconstruction outperforms CPU implementation of the state-of-the art dictionary learning-based approaches by up to two orders of magnitude.",

author = "Quan, \{Tran Minh\} and Jeong, \{Won Ki\}",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2016.",

year = "2016",

doi = "10.1007/978-3-319-46726-9\_56",

language = "English",

isbn = "9783319467252",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "484--492",

editor = "Leo Joskowicz and Sabuncu, \{Mert R.\} and William Wells and Gozde Unal and Sebastian Ourselin",

booktitle = "Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings",

}

TY - GEN

T1 - Compressed sensing dynamic MRI reconstruction using GPU-accelerated 3D convolutional sparse coding

AU - Quan, Tran Minh

AU - Jeong, Won Ki

N1 - Publisher Copyright: © Springer International Publishing AG 2016.

PY - 2016

Y1 - 2016

N2 - In this paper,we introduce a fast alternating method for reconstructing highly undersampled dynamic MRI data using 3D convolutional sparse coding. The proposed solution leverages Fourier Convolution Theorem to accelerate the process of learning a set of 3D filters and iteratively refine the MRI reconstruction based on the sparse codes found subsequently. In contrast to conventional CS methods which exploit the sparsity by applying universal transforms such as wavelet and total variation,our approach extracts and adapts the temporal information directly from the MRI data using compact shift-invariant 3D filters. We provide a highly parallel algorithm with GPU support for efficient computation,and therefore,the reconstruction outperforms CPU implementation of the state-of-the art dictionary learning-based approaches by up to two orders of magnitude.

AB - In this paper,we introduce a fast alternating method for reconstructing highly undersampled dynamic MRI data using 3D convolutional sparse coding. The proposed solution leverages Fourier Convolution Theorem to accelerate the process of learning a set of 3D filters and iteratively refine the MRI reconstruction based on the sparse codes found subsequently. In contrast to conventional CS methods which exploit the sparsity by applying universal transforms such as wavelet and total variation,our approach extracts and adapts the temporal information directly from the MRI data using compact shift-invariant 3D filters. We provide a highly parallel algorithm with GPU support for efficient computation,and therefore,the reconstruction outperforms CPU implementation of the state-of-the art dictionary learning-based approaches by up to two orders of magnitude.

UR - https://www.scopus.com/pages/publications/84996520553

U2 - 10.1007/978-3-319-46726-9_56

DO - 10.1007/978-3-319-46726-9_56

M3 - Conference contribution

AN - SCOPUS:84996520553

SN - 9783319467252

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 484

EP - 492

BT - Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings

A2 - Joskowicz, Leo

A2 - Sabuncu, Mert R.

A2 - Wells, William

A2 - Unal, Gozde

A2 - Ourselin, Sebastian

PB - Springer Verlag

ER -

Compressed sensing dynamic MRI reconstruction using GPU-accelerated 3D convolutional sparse coding

Abstract

Publication series

Bibliographical note

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this