Denoising filters evaluation for magnetic resonance images

Abstract Eleven denoising filters, proposed during last fifteen years, are introduced and compared for magnetic resonance images. Among them, the state-of-art denoising algorithms, NLM and BM3D, have attracted much attention. Several expansions are proposed to improve the noise reduction based on these two algorithms. On the other hand, optimal dictionaries, sparse representations and appropriate shapes of the transform's support are also considered for the image denoising. Based on the estimated noise variance, the comparison of various filters is implemented by measuring the signal-noise-ratio (SNR), resolution and uniformity of a phantom image. The subjective judgment of denoising effectiveness is executed for a clinical image. And the computational time is finally evaluated