Intuitionistic fuzzy sets based credibilistic fuzzy C-means clustering for medical image segmentation

Abstract

Intuitionistic fuzzy c-means (IFCM) is a clustering technique which considers hesitation factor and fuzzy entropy to improve the noise sensitivity of fuzzy c-means (FCM). Credibilistic FCM modified FCM by introducing a term, credibility, to reduce the affect of outliers on the location of cluster centers. In this paper an intutionistic fuzzy set based robust credibilistic IFCM is proposed. Proposed method is tested on real and simulated MRI and CT scan brain images and is compared with seven algorithms namely fuzzy c-means (FCM), Type-2 FCM, IFCM, credibilistic FCM, spatial FCM, possibilistic c-means and probabilistic FCM.