多次扫描相干平均是提高磁共振图像信噪比的常用方法，但如果在多次扫描过程中病人发生自主或不自主的运动，使得图像中的组织发生位移，简单相干平均图像会导致图像模糊.本文受非局域均值算法的启发，提出了一种基于局部位移校正的相干平均方法.该算法通过比较多次采集的图像中组织结构的局部相似性，找出图像间的局部位移，利用该信息修正位移后进行加权平均，从而达到提高图像信噪比的目的.我们用模型及真实的肝脏弥散数据进行了实验.实验结果表明，对于不同次采样间存在运动的磁共振图像，该算法可有效地提高信噪比并保持结构边缘；其结果优于简单的相干平均，去噪效果也优于经典的非局域均值算法.
谢海滨
,
周敏雄
,
向之明
,
李文静
,
严序
,
杨光
. 基于局部位移校正的磁共振图像相干平均[J]. 波谱学杂志, 2017
, 34(3)
: 294
-301
.
DOI: 10.11938/cjmr20162525
In magnetic resonance imaging (MRI), data averaging is often used to improve signal-to-noise ratio (SNR) of the images. However, image blurring can be induced by averaging if movements occur during scanning. Inspired by the patch-matching method used in the non-local means algorithm, a new method to find out local offsets of structures in multiple images was proposed by comparing the neighborhood similarities of the image patches. The local offsets could then be corrected before weighted averaging of the images. The performance of the proposed method was verified with both phantom and patient images. The results demonstrated that the proposed algorithm could improve SNR while preserving the image edges and details correctly.
[1] MOHAN J, KRISHNAVENI V, GUO Y. A survey on the magnetic resonance image denoising methods[J]. Biomed Signal Proces, 2014, 9(1):56-69.
[2] JOSHI N, JAIN S. Optimization of nonlocal means filtering technique for denoising magnetic resonance images:A review[C]//Proceedings of fifth international conference on soft computing for problem solving. Singapore:Springer, 2016:1-15.
[3] SAMSONOV A A, JOHNSON C R. Noise-adaptive nonlinear diffusion filtering of MR images with spatially varying noise levels[J]. Magn Reson Med, 2004, 52(4):798-806.
[4] LU B B, DENG C, LIU Q, et al. Four order adaptive PDE method for MRI denoising[C]//IEEE ACMT Comput Bi, 2009.
[5] ZHANG F, MA L H. MRI denoising using the anisotropic coupled diffusion equations[C]//20103rd international conference on BMEI. IEEE, 2010, 1:397-401.
[6] MA J, PLONKA G. Combined curvelet shrinkage and nonlinear anisotropic diffusion[J]. IEEE T Image Process, 2007, 16(9):2198-2206.
[7] PARTHIBAN L, SUBRAMANIAN R. Medical image denoising using X-lets[C]//India conference, 2006 Annual IEEE, 2006:1-6.
[8] YANG X, FEI B. A wavelet multiscale denoising algorithm for magnetic resonance (MR) images[J]. Meas Sci Technol, 2011, 22(2):025803.
[9] LUISIER F, BLU T, WOLFE P J. A cure for noisy magnetic resonance images:Chi-square unbiased risk estimation[J]. IEEE T Image Process, 2012, 21(8):3454-3466.
[10] BUADES A, COLL B, MOREL J M. A review of image denoising algorithms, with a new one[J]. Multiscale Model Sim, 2005, 4(2):490-530.
[11] MANJÓN J V, CARBONELL J, LULL J J, et al. MRI denoising using non-local means[J]. Med Image Anal, 2008, 12(4):514-523.
[12] HU J R, PU Y F, WU X, et al. Improved DCT-based nonlocal means filter for MR images denoising[J]. Comput Math Methods Med, 2012:232685.
[13] MANJÓN J V, COUPÉ P, BUADES A, et al. New methods for MRI denoising based on sparseness and self-similarity[J]. Med Image Anal, 2012, 16(1):18-27.
[14] ZHANG X Y, HOU G R, MA J H, et al. Denoising MR images using non-local means filter with combined patch and pixel similarity[J]. PloS one, 2014, 9(6):e100240.
[15] GUO T L, LIU Q G, LUO J H. Filter bank based nonlocal means for denoising magnetic resonance images[J]. Journal of Shanghai Jiaotong University (Science), 2014, 19(1):72-78.
[16] WIEST-DAESSLÉ N, PRIMA S, COUPÉ P, et al. Rician noise removal by non-local means filtering for low signal-to-noise ratio MRI:Applications to DT-MRI[M]. Berlin:Springer Heidelberg, 2008:171-179.
[17] BARNES C, SHECHTMAN E, FINKELSTEIN A, et al. PatchMatch:A randomized correspondence algorithm for structural image editing[J]. ACM Graphic, 2009, 28(3):24.
[18] GIRAUD R, TA V T, PAPADAKIS N, et al. An optimized PatchMatch for multi-scale and multi-feature label fusion[J]. Neuroimage, 2016, 124:770-782.
[19] BARNES C, SHECHTMAN E, GOLDMAN D B, et al. The generalized patchmatch correspondence algorithm[M]. Berlin:Springer Heidelberg, 2010:29-43.
[20] WIEST-DAESSLÉ N, PRIMA S, COUPÉ P, et al. Non-local means variants for denoising of diffusion-weighted and diffusion tensor MRI[M]. Berlin:Springer Heidelberg, 2007:344-351.
[21] BAO L, ROBINI M, LIU W, et al. Structure-adaptive sparse denoising for diffusion-tensor MRI[J]. Med Image Anal, 2013, 17(4):442-457.
[22] MANJÓN J V, COUPÉ P, BUADES A, et al. Non-local MRI upsampling[J]. Med Image Anal, 2010, 14(6):784-792.
[23] http://brainweb.bic.mni.mcgill.ca/brainweb/
[24] BUADES A, COLL B, MOREL J M. A non-local algorithm for image denoising[C]. IEEE T Comput, 2005, 2:60-65.
[25] COUPÉ P, YGER P, PRIMA S, et al. An optimized blockwise nonlocal means denoising filter for 3-D magnetic resonance images[J]. IEEE T Med Imaging, 2008, 27(4):425-441.
[26] AVANAKI A N, DIYANAT A, SODAGARI S. Optimum parameter estimation for non-local means image de-noising using corner information[C]. IEEE T Signal Proce, 2008:861-863.
