Denoising of Brain MRI Images Using a Hybrid Filter Method of Sylvester-Lyapunov Equation and Non Local Means

Abstract

Medical imaging is playing an effective role in diagnosing diseases noninvasively. There are many denoising filters, but previous filters have not attained the desirable results. Denoising techniques have to be applied in such a way that, they preserve and enhance diagnostic relevant information of the MRI images. Therefore noise reduction has always been a challenge for researchers. A hybrid denoising technique based on Sylvester-Lyapunov Equation and Non Local Means method has been proposed in this work. This hybrid denoising technique has been compared with the existing techniques of denoising by varying noise levels in brain images of Brainweb dataset and real dataset. Efficiency of methods have been discussed by measuring PSNR and SSIM coefficients.