MRI Denoising Based on a Non-Parametric Bayesian Image Sparse Representation Method

Abstract:

Magnetic Resonance images are often corrupted by Gaussian noise which highly affects the quality of MR images. In this paper, a Non-Parametric hierarchical Bayesian image sparse representation method is proposed to wipe out Gaussian distribution noise coupling in MR images. In this method a spike-slab prior is imposed on sparse coefficients, and a redundant dictionary is learned from the corrupted image. Experimental results show that the method not only improves the effect of MRI denoising, but also can obtain good estimation of the noise variance. Compared to non-local filter method, this model shows better visual quality as well as higher PSNR.

• Periodical: Advanced Materials Research (Volumes 219-220)
• Pages: 1354-1358
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.219-220.1354
• Online since: March 2011
• Authors: Xian Bo Chen, Xing Hao Ding, Hui Liu
• Keywords: Dictionary Learning, MRI Denoising, Non-Parametric Bayesian, Sparse Representation, Spike-Slab Prior

References

[1] B. Vidakovic, Wavelet-based nonparametric Bayes methods,, in Practical Nonparametric and Semiparametric Bayesian Statistics, D. D. Dey,P. Müller, and D. Sinha, Eds. New York: Springer-Verlag, Lecture Notes in Statistics 133, p.133–155, (1998).

DOI: 10.1007/978-1-4612-1732-9_7

[2] A. Pizurica, W. Philips, I. Lemahieu, and M. Acheroy, A versatile wavelet domain noise filtration technique for medical imaging.,IEEE Trans. Med. Imaging, vol. 22, no. 3, p.323–331, (2003).

DOI: 10.1109/tmi.2003.809588

[3] José V. Manjón, José Carbonell-Caballero, Juan J. Lull, Gracián García-Martí, Luís Martí-Bonmatí, Montserrat Robles,MRI denoising using Non-Local Means ,Medical Image Analysis, vol. 12, no. 4, pp.514-523, (2008).

DOI: 10.1016/j.media.2007.03.001

[4] Xiaobo Qu, Xue Cao, Di Guo, Changwei Hu, Zhong Chen,Compressed sensing MRI with combined sparsifying transforms and smoothed l0 norm minimization, ICASSP, 2010 IEEE International Conference ,(2010).

DOI: 10.1109/icassp.2010.5495174

[5] M. Elad and M. Aharon. Image denoising via sparse and redundant representations over learned dictionaries. IEEE Trans. Image Processing, vol.15, no.12, p.3736 – 3745, (2006).

DOI: 10.1109/tip.2006.881969

[6] Mingyuan Zhou, Haojun Chen, John Paisley, Lu Ren, Guillermo Sapiro, and Lawrence Carin, Non-Parametric bayesian dictionary learning for sparse image representations,, in Proc. Neural and Information Processing Systems (NIPS), (2009).

DOI: 10.1109/tip.2011.2160072

[7] Michael E. Tipping,Sparse Bayesian Learning and the Relevance Vector Machine,, journal of Machine Learning Research 1, pp.211-244, (2001).

[8] StCphane G. Mallat and Zhifeng Zhang, Matching Pursuits with Time-Frequency Dictionaries,IEEE Trans on Signal Processing, vol.41, no. 12,pp.3397-3415, (1993).

DOI: 10.1109/78.258082

[9] Chen S , Donoho D , Saunders M. Atomic Decomposition by Basis Pursuit [J ] . SIAM J Sci Comput, vol. 43, no.1, pp.129-159, (2001).

DOI: 10.1137/s003614450037906x

[10] Shihao Ji, Ya Xue, and Lawrence Carin, Bayesian Compressive Sensing, IEEE Trans. Signal Processing, vol. 56, no. 6, pp.2346-2356, (2008).

DOI: 10.1109/tsp.2007.914345