A Neutrosophic approach of MRI denoising

Abstract

In this paper, a new filtering method to remove Rician noise from magnetic resonance image is presented. This filter is based on Neutrosophic set (NS) approach of median filtering. A Neutrosophic set, a part of neutrosophy theory, studies the origin, nature and scope of neutralities, as well as their interactions with different ideational spectra. Here the MRI image is transformed into NS domain, which is described using three membership sets: T, I, F. The entropy of the neutrosophic set is defined and employed to evaluate the indeterminacy. The γ-median filtering operation is used on T and F to decrease the set indeterminacy and remove noise. We have conducted experiments on real MR image with Rician noise added. The performance of this filter is compared with the median filtering and the classical non local mean (NLM) de-noising approaches. This filter outperforms the median filter for different levels of noise. For low signal to noise ratio (SNR), this filter provides high peak signal to noise ratio (PSNR) than the NLM approach.