Identification and removal of different noise patterns by measuring SNR value in magnetic resonance images

In MR image Rician noise is one of the prominent noise, however Gaussian and Rayleigh noise are also present. These types of noises in the MRI can be identified by measuring SNR value of image data. In the literature, there are many methods available to remove Rician noise. But little method has been reported for the removal of Rayleigh and Gaussian noise in MRI. So in this paper we concentrate on removal of Rayleigh and Gaussian noise from MRI. This method is automatically identify various type of noise present into the MRI and filters them by choosing an appropriate filter. The proposed filter consists of two terms namely data fidelity and prior. The data fidelity term i.e. likelihood term is derived from Gaussian pdf and Rayleigh pdf and a nonlinear complex diffusion (CD) based prior is used. The performance analysis and comparative study of the proposed method with other standard methods is presented for Brain Web dataset at varying noise levels in terms of MSE and SSIM. From the simulation results, it is observed that the proposed framework with CD based prior is performing better in comparison to other priors in consideration.