Toward real-time bent-ray breast ultrasound tomography using GPUs

Yassin Labyed; Lianjie Huang

doi:10.1117/12.2043127

Toward real-time bent-ray breast ultrasound tomography using GPUs

Yassin Labyed
, Lianjie Huang^*

^*Corresponding author for this work

Los Alamos National Laboratory

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

11 Scopus citations

Abstract

Breast ultrasound tomography is a promising imaging modality that has the potential to improve the diagnosis and screening of breast cancer. We develop a bent-ray ultrasound tomography algorithm to reconstruct sound-speed images of the breast. We investigate the acceleration of the algorithm using graphical processing units (GPUs). We adapt the algorithmic steps of ultrasound bent-ray tomography to a GPU cluster, and use multi-GPU scaling to speed up the computation. Our results show that it is very promising to use a GPU cluster with multiple GPU cards to achieve nearly real-time tomographic reconstruction.

Original language	English
Title of host publication	Medical Imaging 2014
Subtitle of host publication	Ultrasonic Imaging and Tomography
Publisher	SPIE
ISBN (Print)	9780819498335
DOIs	https://doi.org/10.1117/12.2043127
State	Published - 2014
Externally published	Yes
Event	Medical Imaging 2014: Ultrasonic Imaging and Tomography - San Diego, CA, United States Duration: 18 Feb 2014 → 20 Feb 2014

Publication series

Name	Progress in Biomedical Optics and Imaging - Proceedings of SPIE
Volume	9040
ISSN (Print)	1605-7422

Conference

Conference	Medical Imaging 2014: Ultrasonic Imaging and Tomography
Country/Territory	United States
City	San Diego, CA
Period	18/02/14 → 20/02/14

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

Bent-ray ultrasound tomography
CUDA
Conjugate gradient
General Purpose Processing Units (GPPUs)
Inversion

Access to Document

10.1117/12.2043127

Cite this

@inproceedings{eac70111fbe44effa423a1beb64d9048,

title = "Toward real-time bent-ray breast ultrasound tomography using GPUs",

abstract = "Breast ultrasound tomography is a promising imaging modality that has the potential to improve the diagnosis and screening of breast cancer. We develop a bent-ray ultrasound tomography algorithm to reconstruct sound-speed images of the breast. We investigate the acceleration of the algorithm using graphical processing units (GPUs). We adapt the algorithmic steps of ultrasound bent-ray tomography to a GPU cluster, and use multi-GPU scaling to speed up the computation. Our results show that it is very promising to use a GPU cluster with multiple GPU cards to achieve nearly real-time tomographic reconstruction.",

keywords = "Bent-ray ultrasound tomography, CUDA, Conjugate gradient, General Purpose Processing Units (GPPUs), Inversion",

author = "Yassin Labyed and Lianjie Huang",

year = "2014",

doi = "10.1117/12.2043127",

language = "英语",

isbn = "9780819498335",

series = "Progress in Biomedical Optics and Imaging - Proceedings of SPIE",

publisher = "SPIE",

booktitle = "Medical Imaging 2014",

address = "美国",

note = "Medical Imaging 2014: Ultrasonic Imaging and Tomography ; Conference date: 18-02-2014 Through 20-02-2014",

}

Labyed, Y & Huang, L 2014, Toward real-time bent-ray breast ultrasound tomography using GPUs. in Medical Imaging 2014: Ultrasonic Imaging and Tomography., 90401N, Progress in Biomedical Optics and Imaging - Proceedings of SPIE, vol. 9040, SPIE, Medical Imaging 2014: Ultrasonic Imaging and Tomography, San Diego, CA, United States, 18/02/14. https://doi.org/10.1117/12.2043127

TY - GEN

T1 - Toward real-time bent-ray breast ultrasound tomography using GPUs

AU - Labyed, Yassin

AU - Huang, Lianjie

PY - 2014

Y1 - 2014

N2 - Breast ultrasound tomography is a promising imaging modality that has the potential to improve the diagnosis and screening of breast cancer. We develop a bent-ray ultrasound tomography algorithm to reconstruct sound-speed images of the breast. We investigate the acceleration of the algorithm using graphical processing units (GPUs). We adapt the algorithmic steps of ultrasound bent-ray tomography to a GPU cluster, and use multi-GPU scaling to speed up the computation. Our results show that it is very promising to use a GPU cluster with multiple GPU cards to achieve nearly real-time tomographic reconstruction.

AB - Breast ultrasound tomography is a promising imaging modality that has the potential to improve the diagnosis and screening of breast cancer. We develop a bent-ray ultrasound tomography algorithm to reconstruct sound-speed images of the breast. We investigate the acceleration of the algorithm using graphical processing units (GPUs). We adapt the algorithmic steps of ultrasound bent-ray tomography to a GPU cluster, and use multi-GPU scaling to speed up the computation. Our results show that it is very promising to use a GPU cluster with multiple GPU cards to achieve nearly real-time tomographic reconstruction.

KW - Bent-ray ultrasound tomography

KW - CUDA

KW - Conjugate gradient

KW - General Purpose Processing Units (GPPUs)

KW - Inversion

UR - https://www.scopus.com/pages/publications/84901747238

U2 - 10.1117/12.2043127

DO - 10.1117/12.2043127

M3 - 会议稿件

AN - SCOPUS:84901747238

SN - 9780819498335

T3 - Progress in Biomedical Optics and Imaging - Proceedings of SPIE

BT - Medical Imaging 2014

PB - SPIE

T2 - Medical Imaging 2014: Ultrasonic Imaging and Tomography

Y2 - 18 February 2014 through 20 February 2014

ER -

Toward real-time bent-ray breast ultrasound tomography using GPUs

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this