Medical & Biological Engineering & Computing

, Volume 50, Issue 9, pp 991–1000 | Cite as

Exploiting sparsity and low-rank structure for the recovery of multi-slice breast MRIs with reduced sampling error

  • X. X. Yin
  • B. W.-H. Ng
  • K. Ramamohanarao
  • A. Baghai-Wadji
  • D. AbbottEmail author
Original Article


It has been shown that, magnetic resonance images (MRIs) with sparsity representation in a transformed domain, e.g. spatial finite-differences (FD), or discrete cosine transform (DCT), can be restored from undersampled k-space via applying current compressive sampling theory. The paper presents a model-based method for the restoration of MRIs. The reduced-order model, in which a full-system-response is projected onto a subspace of lower dimensionality, has been used to accelerate image reconstruction by reducing the size of the involved linear system. In this paper, the singular value threshold (SVT) technique is applied as a denoising scheme to reduce and select the model order of the inverse Fourier transform image, and to restore multi-slice breast MRIs that have been compressively sampled in k-space. The restored MRIs with SVT for denoising show reduced sampling errors compared to the direct MRI restoration methods via spatial FD, or DCT. Compressive sampling is a technique for finding sparse solutions to underdetermined linear systems. The sparsity that is implicit in MRIs is to explore the solution to MRI reconstruction after transformation from significantly undersampled k-space. The challenge, however, is that, since some incoherent artifacts result from the random undersampling, noise-like interference is added to the image with sparse representation. These recovery algorithms in the literature are not capable of fully removing the artifacts. It is necessary to introduce a denoising procedure to improve the quality of image recovery. This paper applies a singular value threshold algorithm to reduce the model order of image basis functions, which allows further improvement of the quality of image reconstruction with removal of noise artifacts. The principle of the denoising scheme is to reconstruct the sparse MRI matrices optimally with a lower rank via selecting smaller number of dominant singular values. The singular value threshold algorithm is performed by minimizing the nuclear norm of difference between the sampled image and the recovered image. It has been illustrated that this algorithm improves the ability of previous image reconstruction algorithms to remove noise artifacts while significantly improving the quality of MRI recovery.


Singular value threshold Model order reduction Sparse sensing MRI k-space Low rank matrix Finite differences Discrete cosine transform 



The MRI datasets were afforded by Q. Yang, Apollo Medical Imaging Technology Pty. Ltd. North Melbourne, VIC 3051, Australia, and A. Pitman, Department of Anatomy and Cell Biology, The University of Melbourne and Sydney School of Medicine, The University of Notre Dame, Australia. This work was supported in part by the Australian Research Council (ARC) Discovery Project funding scheme—Project No. DP0988064. This project was also carried out under an Australian Research Council (ARC) Linkage Grant—LP0775463.


  1. 1.
    Lustig M, Donoho D, Pauly JM (2007) Sparse MRI: the application of compressed sensing for rapid MR imaging. Magn Reson Med 58(6):1182–1195PubMedCrossRefGoogle Scholar
  2. 2.
    Ljunggren S (1983) A simple graphical representation of fourier-based imaging methods. J Magn Reson 54(2):338–343Google Scholar
  3. 3.
    Twieg D (1983) The k-trajectory formulation of the NMR imaging process with applications in analysis and synthesis of imaging methods. Med Phys 10(5):610–621PubMedCrossRefGoogle Scholar
  4. 4.
    Tsaig Y, Donoho D (2006) Extensions of compressed sensing. Signal Process 86(3):533–548CrossRefGoogle Scholar
  5. 5.
    Chralampidis D, Kasparis T, Georgiopoulos M (2001) Classification of noisy signals using fuzzy ARTMAP neural networks. IEEE Trans Neural Netw 12(5):1023–1036PubMedCrossRefGoogle Scholar
  6. 6.
    Wichmann F, Braun D, Gegenfurtner K (2006) Phase noise and the classification of natural images. Vis Res 46(8–9):1520–1529PubMedCrossRefGoogle Scholar
  7. 7.
    Wiener N (1949) Extrapolation, interpolation, and smoothing of stationary time series. Wiley, New YorkGoogle Scholar
  8. 8.
    Cai JF, Candès EJ, Shen Z (2010) A singular value thresholding algorithm for matrix completion’. SIAM J Optim 20(4):1956–1982CrossRefGoogle Scholar
  9. 9.
    Lingala S, Yue H, DiBella E, Jacob M (2011) Accelerated dynamic MRI exploiting sparsity and low-rank structure: k-t SLR’. IEEE Trans Med Imaging 30(5):1042–1054PubMedCrossRefGoogle Scholar
  10. 10.
    Recht B, Fazel M, Parrilo PA (2010) Guaranteed minimum-rank solutions of linear matrix equations via nuclear norm minimization. SIAM Rev 52(3):471–501CrossRefGoogle Scholar
  11. 11.
    Osher S, Mao Y, Dong B, Yin W (2010) Fast linearized Bregman iteration for compressive sensing and sparse denoising. Commun Math Sci 8(1):93–111Google Scholar
  12. 12.
    Antoulas A (2005) Approximation of large-scale dynamical systems. SIAM, New YorkCrossRefGoogle Scholar
  13. 13.
    Golub GH, Van Loan CF (1996) Matrix computations. The John Hopkins University Press, Baltimore, MDGoogle Scholar
  14. 14.
    Schneider MK, Willsky A (2001) Krylov subspace estimation. SIAM J Sci Comput 22(5):1840–1864CrossRefGoogle Scholar
  15. 15.
    Bai Z (2002) Krylov subspace techniques for reduced-order modeling of large-scale dynamical systems. Appl Num Math 43(1–2):9–44CrossRefGoogle Scholar
  16. 16.
    Rudin L, Osher S, Fatemi E (1992) Non-linear total variation noise removal algorithm. Phys D 60:259–268CrossRefGoogle Scholar
  17. 17.
    Huynh-Thu Q, Ghanbari M (2008) Scope of validity of PSNR in image/video quality assessment. Electron Lett 44:800–801CrossRefGoogle Scholar
  18. 18.
    Mu T, Nandi AK, Rangayyan RM (2007) Classification of breast masses via nonlinear transformation of features based on a kernel matrix. Med Biol Eng Comput 45(8):769–780PubMedCrossRefGoogle Scholar
  19. 19.
    Sun F, Morris D, Babyn P (2009) The optimal linear transformation-based fMRI feature space analysis. Med Biol Eng Comput 47(11):1119–1129PubMedCrossRefGoogle Scholar
  20. 20.
    Yin XX, Ng BW-H, Yang Q, Pitman A, Ramamohanarao K, Abbott D (2012) Anatomical landmark localization in breast dynamic contrast-enhanced MR imaging. Med Biol Eng Comput 50(1):91–101PubMedCrossRefGoogle Scholar

Copyright information

© International Federation for Medical and Biological Engineering 2012

Authors and Affiliations

  • X. X. Yin
    • 1
    • 2
    • 3
  • B. W.-H. Ng
    • 1
  • K. Ramamohanarao
    • 2
  • A. Baghai-Wadji
    • 3
  • D. Abbott
    • 1
    Email author
  1. 1.Centre for Biomedical Engineering, School of Electrical and Electronic EngineeringThe University of AdelaideAdelaideAustralia
  2. 2.Department of Computer Science and Software Engineering, The Melbourne School of EngineeringThe University of MelbourneMelbourneAustralia
  3. 3.School of Electrical and Computer Engineering, Royal Melbourne Institute of TechnologyMelbourneAustralia

Personalised recommendations