MRI Brain Tumour Segmentation Based on Multimodal Clustering and Level-set Method

The process of partitioning an image into mutually exclusive regions is called image segmentation. Brain tumour segmentation is still a challenging task in MRI imaging. This paper presents a simple and efficient brain tumour segmentation approach, aiming to extract the whole tumour, based on multimodal clustering and level-set method. Two clustering techniques were used namely: Particle Swarm optimization (PSO) and Fuzzy c-mean (FCM). The Brain Tumour Segmentation database (BRATS) 2013 was used for the evaluation. For comparison, results of the common single-modal clustering using PSO and FCM with level-set were presented. The results revealed that the proposed method using multimodal PSO clustering is the best approach compared to single-modal clustering of PSO and FCM, or multimodal FCM.