Diffusion Weighted MRI Rician Noise Restoration Using Modified Wiener Filtering
1 Department of Electronic Engineering, Chengdu University of Information Technology, Chengdu 610225, China
2 College of Computer Science, Sichuan University, Chengdu 610065, China
Abstract：According to Rician noise distribution from diffusion weighted magnetic resonance image and the bias from the traditional Wiener filter which is designed for Gaussian model, this paper encapsulated multi diffusion weighted magnetic resonance images from nearby directions for Wiener filtering. In the procedure, Wiener filter was modified for Rician noise model, and the parameters of filter were estimated through anisotropic area for further improvement of restoration. The simulation and experiment of both synthetic and in vivo diffusion weighted magnetic resonance image data demonstrated that the proposed method can effectively remove the noise in the diffusion weighted magnetic resonance image, and improve the quality and orientation information of diffusion tensor MRI. In 10% Rician noise condition, peak signal noise ratio of diffusion weighted magnetic resonance image was increased 10 dB, and mean angular variation of diffusion tensor MRI decreased 5 degree, which ensures better accuracy and robustness of the further applications.
［1］Basser P, Mattiello J, LeBihan D. MR diffusion tensor spectroscopy and imaging ［J］. Biophysical Journal, 1994, 66（1）: 259-267.
［2］Lazar M. Mapping brain anatomical connectivity using white matter tractography ［J］. NMR in Biomedicine. 2010, 23: 821-835.
［3］吴锡，周激流，谢明元, 等. 磁共振弥散张量成像的脑白质纤维优化重建［J］. 中国生物医学工程学报，2009, 28(6):931-934。
［4］Heemskerk AM, Sinha TK, Wilson KJ, et al. Repeatability of DTIbased muscle fiber tracking ［J］. NMR in Biomedicine, 2010, 23:294-303.
［5］Savannah CP, Revathi SM, Ali Z, et al.Diffusion tensor magnetic resonance imaging of the normal breast ［J］. Magnetic Resonance Imaging, 2010, 28:320-328.
［6］Anderson AW. Theoretical analysis of the effects of noise on diffusion tensor imaging ［J］. Magnetic Resonance in Medicine, 2001, 46(6):1174-1188.
［7］Perona P, Malik J. Scalespace and edge detection using anisotropic diffusion ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1990, 12： 629-639.
［8］Parker GJ, Schnabel JA, Symms MR, et al. Nonlinear smoothing for reduction of systematic and random errors in diffusion tensor imaging ［J］. J Magn Reson Imaging, 2000,11:702-710.
［9］McGraw T, Vemuri BC, Chen Ying, et al. DTMRI denoising and neural fiber tracking ［J］. Medical Image Analysis, 2004, 8: 95-111.
［10］Sijbers J, den Dekker AJ. Maximum Likelihood estimation of signal amplitude and noise variance from MR data ［J］. Magnetic Resonance Imaging, 2004, 51: 586-594.
［11］Pizurica A, Philips W, Lemahieu I, et al.A versatile Wavelet domain noise filtration technique for medical imaging ［J］. IEEE Transactions on Medical Imaging, 2003, 22 (3): 323-331.
［12］Aja-Fernández S, Niethammer M, Kubicki M, et al.Restoration of DWI data using a Rician LMMSE estimator ［J］. IEEE Transactions on Medical Imaging, 2008, 27 (10), 1389-1403.
［13］白衡，高玉蕊，王世杰，等.DTI扩散张量的一种稳健估计方法［J］. 计算机研究与发展， 2008，45（7）：1232-1238.
［14］Ding Z, Gore JC, Anderson AW. Reduction of noise in diffusion tensor images using anisotropic smoothing ［J］. Magnetic Resonance in Medicine, 2005, 53: 485-490.
［15］RuizAlzola J, AlberolaLopez C, Westin CF. Kriging filters for multidimensional signal processing ［J］. Signal Processing, 2005, 85: 413-439.
［16］AjaFernández S, AlberolaLópez C, Westin CF. Noise and signal estimation in magnitude MRI and Rician distributed images: a LMMSE approach ［J］. IEEE Transactions on Image Processing, 2008, 17 (8): 1383–1398.
［17］Hillery AD, Chin RT. Iterative Wiener filters for image restoration ［J］. IEEE Transactions on Signal Processing, 1991,39: 1892-1899.
