Parallel implementation of L + S signal recovery in dynamic MRI

Abstract

Dynamic MRI is useful to diagnose different diseases, e.g. cardiac ailments, by monitoring the structure and function of the heart and blood flow through the valves. Faster data acquisition is highly desirable in dynamic MRI, but this may lead to aliasing artifacts due to under-sampling. Advanced image reconstruction algorithms are required to obtain aliasing-free MR images from the acquired under-sampled data. One major limitation of using the advanced reconstruction algorithms is their computationally expensive and time-consuming nature, which make them infeasible for clinical use, especially for applications like cardiac MRI. L + S decomposition model is an approach provided in literature which separates the sparse and low-rank information in dynamic MRI. However, L + S decomposition model is a computationally complex process demanding significant computation time. In this paper, a parallel framework is proposed to accelerate the image reconstruction process of L + S decomposition model using GPU. Experiments are performed on cardiac perfusion dataset (\(256\, \times \,256\, \times \,40\, \times \,12\)) and cardiac cine dataset (\(256\, \times \,256\, \times \,11\, \times \,30\)) using NVIDIA’s GeForce GTX780 GPU and Core-i7 CPU. The results show that the proposed method provides up to 18 × speed-up including the memory transfer time (i.e. data transfer between the CPU and GPU) and ~ 46 × speed-up without memory transfer for the cardiac perfusion dataset in our experiments. This level of improvement in the reconstruction time will increase the usefulness of L + S reconstruction by making it feasible for clinical applications.

This is a preview of subscription content, log in to check access.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9

References

  1. 1.

    McRobbie DW, Moore EA, Graves MJ, Prince MR (2007) MRI from Picture to Proton. Cambridge University Press, Cambridge, pp 51–57

    Google Scholar 

  2. 2.

    Oakes RS, Badger TJ, Kholmovski EG, Akoum N, Burgon NS, Fish EN, Daccarett M (2009) Detection and quantification of left atrial structural remodeling with delayed-enhancement magnetic resonance imaging in patients with atrial fibrillation. Circulation 119:1758–1767

    Article  Google Scholar 

  3. 3.

    Alonzi R, Padhani AR, Allen C (2007) Dynamic contrast enhanced MRI in prostate cancer. Eur J Radiol 63(3):335–350

    Article  Google Scholar 

  4. 4.

    Manning HC (2011) Characterizing tissue properties with exogenous contrast agents. In: Quantitative MRI in Cancer, vol 1, pp 125–133

  5. 5.

    Gamper U, Boesiger P, Kozerke S (2008) Compressed sensing in dynamic MRI. Magn Reson Med 59(2):365–373

    Article  Google Scholar 

  6. 6.

    Lustig M, Donoho DL, Santos JM, Pauly JM (2008) Compressed sensing MRI. IEEE Signal Process Mag 25(2):72–82

    Article  Google Scholar 

  7. 7.

    Lustig M, Donoho D, Pauly JM (2007) Sparse MRI: The application of compressed sensing for rapid MR imaging. Magn Reson Med 58(6):1182–1195

    Article  Google Scholar 

  8. 8.

    Otazo R, Candès E, Sodickson DK (2015) Low-rank plus sparse matrix decomposition for accelerated dynamic MRI with separation of background and dynamic components. Magn Reson Med 73(3):1125–1136

    Article  Google Scholar 

  9. 9.

    Ravishankar S, Moore BE, Nadakuditi RR, Fessler JA (2017) Low-rank and adaptive sparse signal (LASSI) models for highly accelerated dynamic imaging. IEEE Trans Med Imaging 36(5):1116–1128

    Article  Google Scholar 

  10. 10.

    Roohi SF, Zonoobi D, Kassim AA, Jaremko JL (2017) Multi-dimensional low rank plus sparse decomposition for reconstruction of under-sampled dynamic MRI. Pattern Recogn 63:667–679

    Article  Google Scholar 

  11. 11.

    Liu Q, Wang S, Liang D (2017) Sparse and dense hybrid representation via subspace modeling for dynamic MRI. Comput Med Imaging Graph 56:24–37

    Article  Google Scholar 

  12. 12.

    Owens JD, Houston M, Luebke D, Green S, Stone JE, Phillips JC (2008) GPU computing. Proc IEEE 96(5):879–899

    Article  Google Scholar 

  13. 13.

    Stone SS, Haldar JP, Tsao SC, Sutton BP, Liang ZP (2008) Accelerating advanced MRI reconstructions on GPUs. J Parallel Distrib Comput 68(10):1307–1318

    CAS  Article  Google Scholar 

  14. 14.

    Feng C, Zhao D, Huang M (2016) Image segmentation using CUDA accelerated non-local means denoising and bias correction embedded fuzzy c-means (BCEFCM). Signal Process 122:164–189

    Article  Google Scholar 

  15. 15.

    Feng C, Zhao D (2015) CUDA accelerated uniform re-sampling for non-Cartesian MR reconstruction. Bio-Med Mater Eng 26(s1):S983–S989

    Article  Google Scholar 

  16. 16.

    Feng C, Yang J, Zhao D, Liu J (2013) CUDA accelerated method for motion correction in MR PROPELLER imaging. Magn Reson Imaging 31(8):1390–1398

    Article  Google Scholar 

  17. 17.

    Yang J, Feng C, Zhao D (2013) A CUDA-based reverse gridding algorithm for MR reconstruction. Magn Reson Imaging 31(2):313–323

    Article  Google Scholar 

  18. 18.

    Stone JE, Hallock MJ, Phillips JC, Peterson JR, Luthey-Schulten Z, Schulten K (2016) Evaluation of emerging energy-efficient heterogeneous computing platforms for biomolecular and cellular simulation workloads. In: IEEE International Parallel and Distributed Processing Symposium Workshops, IEEE, pp. 89–100.

  19. 19.

    Smith DS, Gore JC, Yankeelov TE, Welch EB (2012) Real-time compressive sensing MRI reconstruction using GPU computing and split Bregman methods. Int J Biomed Imaging. https://doi.org/10.1155/2012/864827

    Article  PubMed  PubMed Central  Google Scholar 

  20. 20.

    Quan TM, Jeong WK (2016) Compressed sensing dynamic MRI reconstruction using GPU-accelerated 3D convolutional sparse coding. International conference on medical image computing and computer-assisted intervention. Springer International Publishing, Berlin, pp 484–492

    Google Scholar 

  21. 21.

    Shahzad H, Sadaqat MF, Hassan B, Abbasi W, Omer H (2016) Parallel MRI Reconstruction Algorithm Implementation on GPU. Appl Magn Reson 47(1):53–61

    Article  Google Scholar 

  22. 22.

    Tao S, Trzasko JD, Shu Y, Huston J, Bernstein MA (2015) Integrated image reconstruction and gradient nonlinearity correction. Magn Reson Med 74(4):1019–1031

    Article  Google Scholar 

  23. 23.

    Hansen MS, Atkinson D, Sorensen TS (2008) Cartesian SENSE and k-t SENSE reconstruction using commodity graphics hardware. Magnet Reson Med 59(3):463–468

    Article  Google Scholar 

  24. 24.

    Ullah I, Nisar H, Raza H, Qasim M, Inam O, Omer H (2018) QR-decomposition based SENSE reconstruction using parallel architecture. Comput Biol Med 1(95):1–2

    Article  Google Scholar 

  25. 25.

    Sohaib AQ, Ullah I, Omer H (2016) Implementation of Low-Rank + Sparse matrix decomposition on GPUs for accelerating Re-construction time. In: 33rd Annual Scientific Meeting of ESMRMB, Vienna, Austria.

  26. 26.

    Gao H, Cai JF, Shen Z, Zhao H (2011) Robust principal component analysis-based four-dimensional computed tomography. Phys Med Biol 56(11):3181

    Article  Google Scholar 

  27. 27.

    Chandrakasan AP, Sheng S, Brodersen RW (1992) Low-power CMOS digital design. IEICE Trans Electron 75(4):371–382

    Google Scholar 

  28. 28.

    Geer D (2005) Chip makers turn to multicore processors. Computer 38(5):11–13

    Article  Google Scholar 

  29. 29.

    Parallel Computing Toolbox (2017) https://www.mathworks.com/help/distcomp/. Accessed 15 Jan 2017.

  30. 30.

    Ji JX, Son JB, Rane SD (2007) PULSAR: a Matlab toolbox for parallel magnetic resonance imaging using array coils and multiple channel receivers. Concepts Magnet Reson B 31(1):24–36

    Article  Google Scholar 

  31. 31.

    Voronin S, Chartrand R (2013) A new generalized thresholding algorithm for inverse problems with sparsity constraints. In: IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 1636–1640.

  32. 32.

    Seiberlich N, Breuer F, Heidemann R, Blaimer M, Griswold M, Jakob P (2008) Reconstruction of undersampled non-Cartesian data sets using pseudo-CartesianGRAPPA in conjunction with GROG. Magn Reson Med 59(5):1127–1137

    Article  Google Scholar 

  33. 33.

    GPU Computing (2017) https://www.mathworks.com/help/distcomp/gpu-computing.html. Accessed 1 Jan 2017.

Download references

Author information

Affiliations

Authors

Corresponding author

Correspondence to Sohaib A. Qazi.

Ethics declarations

Conflict of interest

Sohaib Ayyaz Qazi declares that he has no conflict of interest. Fareena Tariq declares that he has no conflict of interest. Irfan Ullah declares that he has no conflict of interest. Hammad Omer declares that he has no conflict of interest.

Ethical approval

All the procedures performed in the studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed consent

Informed consent was obtained from all the individual participants included in the study before data acquisition.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Qazi, S.A., Tariq, F., Ullah, I. et al. Parallel implementation of L + S signal recovery in dynamic MRI. Magn Reson Mater Phy (2020). https://doi.org/10.1007/s10334-020-00861-5

Download citation

Keywords

  • Cardiac MRI
  • GPU computing
  • MRI
  • Reconstruction
  • CUDA
  • Compressed sensing