A picture fuzzy clustering approach for brain tumor segmentation

This paper presents a Picture Fuzzy Clustering (PFC) method for MRI brain image segmentation. The PFC is based on the Picture fuzzy set, which is the generalization of the traditional fuzzy set and intuitionistic fuzzy set. In traditional fuzzy set, the problem of uncertainty arises in defining the membership function. Intuitionistic fuzzy set handles this uncertainty by considering hesitation degree. However, intuitionistic fuzzy set fails to solve real time problems which require answers like yes, abstain, no and refusal. The picture fuzzy set solves these problems by considering refusal degree along with membership, neutral and nonmembership degree. Thus, the cluster centers in the PFC may converge to a desirable location than the cluster centers obtained using traditional Fuzzy C-Means (FCM) and Intuitionistic Fuzzy Clustering (IFC). Experimentation is carried out on the standard MRI brain image dataset. To assess the performance, the proposed method is compared with the existing FCM and IFC methods. Results show that the proposed method gives the better result.