Abstract

Computational acceleration on graphics processing units (GPUs) can make advanced magnetic resonance imaging (MRI) reconstruction algorithms attractive in clinical settings, thereby improving the quality of MR images across a broad spectrum of applications. This paper describes the acceleration of such an algorithm on NVIDIA's Quadro FX 5600. The reconstruction of a 3D image with 1283 voxels achieves up to 180 GFLOPS and requires just over one minute on the Quadro, while reconstruction on a quad-core CPU is twenty-one times slower. Furthermore, for the data set studied in this article, the percent error exhibited by the advanced reconstruction is roughly three times lower than the percent error incurred by conventional reconstruction techniques.

Original languageEnglish (US)
Pages (from-to)1307-1318
Number of pages12
JournalJournal of Parallel and Distributed Computing
Volume68
Issue number10
DOIs
StatePublished - Oct 1 2008

Fingerprint

Magnetic Resonance Imaging
Graphics Processing Unit
Magnetic resonance
Imaging techniques
Percent
Program processors
Voxel
Reconstruction Algorithm
3D Image
Graphics processing unit

Keywords

  • CUDA
  • GPU computing
  • MRI
  • Reconstruction

ASJC Scopus subject areas

  • Computer Science Applications
  • Hardware and Architecture
  • Control and Systems Engineering

Cite this

Accelerating advanced MRI reconstructions on GPUs. / Stone, S. S.; Haldar, J. P.; Tsao, S. C.; Hwu, W. m W; Sutton, B. P.; Liang, Z. P.

In: Journal of Parallel and Distributed Computing, Vol. 68, No. 10, 01.10.2008, p. 1307-1318.

Research output: Contribution to journalArticle

@article{f450d4f5a7e74815bc055ff4b546a116,
title = "Accelerating advanced MRI reconstructions on GPUs",
abstract = "Computational acceleration on graphics processing units (GPUs) can make advanced magnetic resonance imaging (MRI) reconstruction algorithms attractive in clinical settings, thereby improving the quality of MR images across a broad spectrum of applications. This paper describes the acceleration of such an algorithm on NVIDIA's Quadro FX 5600. The reconstruction of a 3D image with 1283 voxels achieves up to 180 GFLOPS and requires just over one minute on the Quadro, while reconstruction on a quad-core CPU is twenty-one times slower. Furthermore, for the data set studied in this article, the percent error exhibited by the advanced reconstruction is roughly three times lower than the percent error incurred by conventional reconstruction techniques.",
keywords = "CUDA, GPU computing, MRI, Reconstruction",
author = "Stone, {S. S.} and Haldar, {J. P.} and Tsao, {S. C.} and Hwu, {W. m W} and Sutton, {B. P.} and Liang, {Z. P.}",
year = "2008",
month = "10",
day = "1",
doi = "10.1016/j.jpdc.2008.05.013",
language = "English (US)",
volume = "68",
pages = "1307--1318",
journal = "Journal of Parallel and Distributed Computing",
issn = "0743-7315",
publisher = "Academic Press Inc.",
number = "10",

}

TY - JOUR

T1 - Accelerating advanced MRI reconstructions on GPUs

AU - Stone, S. S.

AU - Haldar, J. P.

AU - Tsao, S. C.

AU - Hwu, W. m W

AU - Sutton, B. P.

AU - Liang, Z. P.

PY - 2008/10/1

Y1 - 2008/10/1

N2 - Computational acceleration on graphics processing units (GPUs) can make advanced magnetic resonance imaging (MRI) reconstruction algorithms attractive in clinical settings, thereby improving the quality of MR images across a broad spectrum of applications. This paper describes the acceleration of such an algorithm on NVIDIA's Quadro FX 5600. The reconstruction of a 3D image with 1283 voxels achieves up to 180 GFLOPS and requires just over one minute on the Quadro, while reconstruction on a quad-core CPU is twenty-one times slower. Furthermore, for the data set studied in this article, the percent error exhibited by the advanced reconstruction is roughly three times lower than the percent error incurred by conventional reconstruction techniques.

AB - Computational acceleration on graphics processing units (GPUs) can make advanced magnetic resonance imaging (MRI) reconstruction algorithms attractive in clinical settings, thereby improving the quality of MR images across a broad spectrum of applications. This paper describes the acceleration of such an algorithm on NVIDIA's Quadro FX 5600. The reconstruction of a 3D image with 1283 voxels achieves up to 180 GFLOPS and requires just over one minute on the Quadro, while reconstruction on a quad-core CPU is twenty-one times slower. Furthermore, for the data set studied in this article, the percent error exhibited by the advanced reconstruction is roughly three times lower than the percent error incurred by conventional reconstruction techniques.

KW - CUDA

KW - GPU computing

KW - MRI

KW - Reconstruction

UR - http://www.scopus.com/inward/record.url?scp=51449100575&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=51449100575&partnerID=8YFLogxK

U2 - 10.1016/j.jpdc.2008.05.013

DO - 10.1016/j.jpdc.2008.05.013

M3 - Article

AN - SCOPUS:51449100575

VL - 68

SP - 1307

EP - 1318

JO - Journal of Parallel and Distributed Computing

JF - Journal of Parallel and Distributed Computing

SN - 0743-7315

IS - 10

ER -