Identification of Brain Tumor from MRI Images using Segmentation

Abstract

A brain tumor is one of the hazardous diseases. Hence, it is necessary to detect the brain tumor appropriately. The detection of brain tumor by processing of magnetic resonance image (MRI) is one of the medical imaging techniques. It is most difficult task for clinical experts to extract the brain tumor by manual inspection because there is variation in tumor size, shape and location. Hence, image processing technique is used for detection of brain tumor. In this paper, by using K means clustering method brain MRI image is segregated in clusters. Then the cluster with tumor is found. But noise can degrade the quality of the image and it can affect the performance of segmentation process. Hence, anisotropic diffusion filter is used for noise reduction, before segmentation process. Morphological operations are applied to segmented image for image enhancement. Finally, the tumor is detected. Some useful features are extracted using gray- level co-occurrence matrix.