Abstract
Magnetic resonance imaging (MRI) is a safe, non-ionizing and powerful diagnostic imaging modality and has a large number of variable contrast mechanisms. There is a fundamental limit in MRI data collection time which can be overcome by using parallel imaging algorithms, e.g., SENSE. Graphical processing units (GPUs) using compute unified device architecture have great potential to reduce the scan time by exploiting the inherent parallelism present in parallel imaging algorithms for MR image reconstruction. This work implements SENSE algorithm using GPU and compares the results with multi-core CPU implementation of SENSE. The inversion of the encoding matrix (formed from the under-sampled data) is a key process in SENSE. The encoding matrix is usually rectangular because the number of receiver coils need to be greater than the acceleration factor. This paper implements the inversion of the rectangular matrix on GPU using Left Inverse Method. All the scripts are written by the authors for this implementation of SENSE on GPU. The results show that GPU attains approximately 7× ~ 28× reduction in SENSE reconstruction time as compared to CPU while maintaining the image quality.
Similar content being viewed by others
References
K.P. Pruessmann, W. Markus, B.S. Markus, B. Peter, Magn. Reson. Med. 42, 952–962 (1999)
M.A. Griswold, P.M. Jakob, R.M. Heidemann, M. Nittka, V. Jellus, J. Wang, B. Kiefer, A. Haase, Magn. Reson. Med. 47(6), 1202–1210 (2002)
J.L. David, Phys. Med. Biol. 52, R15–R55 (2007)
B.K. David, W.H. Wen-mei, Programming Massively Parallel Processors: a Hands-on Approach (Morgan Kaufmann Publishers, USA, 2010)
E. Anders, D. Paul, F. Daniel, M.L. Stephen, Med. Image Anal. 17, 1073–1094 (2013)
X. Lei, Med. Phys. 38, 2685–2697 (2011)
S.S. Stone, J.P. Haldar, S.C. Tsao, W.W. Hwu, B.P. Sutton, Z.-P. Liang, J. Parallel Distrib. Comput. 68(10), 1307–1318 (2008)
T. Schiwietz, T. Chang, P. Speier, R. Westermann, in Proceedings of SPIE, Medical Imaging 2006: Physics of Medical Imaging, vol. 6142 (2006). doi:10.1117/12.652223
S.H. Michael, A. David, S.S. Thomas, Magn. Reson. Med. 59, 463–468 (2008)
S.S. Thomas, A. David, S. Tobias, S.H. Micheal, IEEE Trans. Med. Imaging 28(12) (2009)
MIT OpenCourseWare, “Left and right inverses; pseudoinverse,” [Online]. http://ocw.mit.edu/courses/mathematics/18-06sc-linear-algebra-fall-2011/positive-definite-matrices-and-applications/left-and-right-inverses-pseudoinverse/MIT18_06SCF11_Ses3.8sum.pdf. Accessed March 2015
O. Hammad, D. Robert, Concepts Magn. Reson. Part A 38A, 52–60 (2011)
Nvidea Corp., NVIDIA CUDA TOOLKIT V6.5. August 2014
Ivan. How to measure time in NVIDEA CUDA. Retrieved May 2015 from Ivan’s blog: https://ivanlife.wordpress.com/2011/05/09/time-cuda/. Accessed May 2011
Harris, M. How to implement performance metrics in CUDA C/C++. Retrieved May 2015, from NVIDEA CUDA Zone. http://devblogs.nvidia.com/parallelforall/how-implement-performance-metrics-cuda-cc. Accessed Nov 2012
Microsoft Co. GetTickCount function. Retrieved August 2015, from MSDN Library: https://msdn.microsoft.com/en-us/library/windows/desktop/ms724408(v=vs.85).aspx. Accessed 2009
O. Hammad, R. Dickinson, Concepts Magn. Reson. Part A 36A(3), 178–186 (2010)
P.M. Robson, A.K. Grant, A.J. Madhuranthakam, R. Lattanzi, D.K. Sodickson, C.A. McKenzie, Magn. Reson. Med. 60(3), 895–907 (2008)
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Shahzad, H., Sadaqat, M.F., Hassan, B. et al. Parallel MRI Reconstruction Algorithm Implementation on GPU. Appl Magn Reson 47, 53–61 (2016). https://doi.org/10.1007/s00723-015-0728-6
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00723-015-0728-6