Denoising is always a challenging problem in magnetic resonance imaging (MRI) and is important for clinical\r\ndiagnosis and computerized analysis, such as tissue classification and segmentation. The noise in MRI has a Rician\r\ndistribution. Unlike additive Gaussian noise, Rician noise is signal dependent, and separating the signal from the\r\nnoise is a difficult task. In this paper, we propose a useful alternative of the nonlocal mean (NLM) filter that uses\r\nnonparametric principal component analysis (NPCA) for Rician noise reduction in MR images. This alternative is\r\ncalled the NPCA-NLM filter, and it results in improved accuracy and computational performance. We present an\r\napplicable method for estimating smoothing kernel width parameters for a much larger set of images and\r\ndemonstrate that the number of principal components for NPCA is robust to variations in the noise as well as in\r\nimages. Finally, we investigate the performance of the proposed filter with the standard NLM filter and the PCANLM\r\nfilter on MR images corrupted with various levels of Rician noise. The experimental results indicate that the\r\nNPCA-NLM filter is the most robust to variations in images, and shows good performance at all noise levels tested.
Loading....