A novel fuzzy Dempster–Shafer inference system for brain MRI segmentation

Brain Magnetic Resonance Imaging (MRI) segmentation is a challenging task due to the complex anatomical structure of brain tissues as well as intensity non-uniformity, partial volume effects and noise. Segmentation methods based on fuzzy approaches have been developed to overcome the uncertainty caused by these effects. In this study, a novel combination of fuzzy inference system and Dempster–Shafer Theory is applied to brain MRI for the purpose of segmentation where the pixel intensity and the spatial information are used as features. In the proposed modeling, the consequent part of rules is a Dempster–Shafer belief structure. The novelty aspect of this work is that the rules are paraphrased as evidences. The results show that the proposed algorithm, called FDSIS has satisfactory outputs on both simulated and real brain MRI datasets.