Brain Tumor Segmentation on MR Image Using K-Means and Fuzzy-Possibilistic Clustering

Automated tumor segmentation and estimation from the magnetic resonance imaging (MRI) is a very crucial task from medical point of view due to high varieties of tumor tissues. The advantage of using the MR images is to provide the anatomical structure of the brain that plays a significant role during automated brain tumor detection. In this work, a method for brain tumor segmentation from MR images is proposed which is based on fuzzy-possibilistic C-means (FPCM) and shape based topological properties to identify the exact tumor region. A patch based K-means method is also implemented for skull stripping (brain tissue extraction) as a preprocessing step. Experimental results show that the proposed method has achieved better performance based on volume metrics than previous state-of-the-art algorithms with respect to ground truth (manual segmentation) on MRI standard benchmark datasets.