Feature-Based Segmentation of Brain Tumor in MRI Images

Image segmentation of magnetic resonance image (MRI) is a crucial process for visualization and examination of abnormal tissues, especially during clinical analysis. This paper presents an automatic lesion recognition method in MRI using image segmentation. In the proposed method, the intent region initialization is performed using low-level information by keypoint descriptor and the set of features and filter training data are analyzed to track tumor region. Further region growing and refinement process identifies consist boundary of tumors. The experimental outcomes on BrainWeb and BRATS 2013 datasets validate the effectiveness of the proposed method as that of existing techniques.