Lesion Identification and Tissue Segmentation in Magnetic Resonance (MR) Image using Interval type based Clustering

Abstract— Identifying the lesions and tissue segmentation is a difficult task that helps early diagnosis in radiotherapy. For the several years, different types of segmentation techniques have been proposed but still it is a hard task to identify the lesions. To solve this issues, we introduced the new automated techniques for identify the various types of lesions and to segment the tissues in MR image using interval type based clustering. The interval type based clustering will better results will provide results for solving the optimization problem. Feasible solutions are provided by identifying the optimal cluster; it will improve the accuracy of segmentation process. Different types of lesions and tissue structure present in human brain were segmented will give better visualization for the radiologist. In order to prove the accuracy of the algorithm, the comparison parameters such as Dice overlap index (DOI), Computation time are used.


Index Terms— lesions identification, Interval type based clustering, Magnetic Resonance Image (MRI)