Automatic Detection of Brain Tumor Using K-Means Clustering

Abstract

Brain tumor is an uncommon and uncontrolled growth of cell in brain. In medical image processing one of the most challenging tasks is study of Brain tumor. Recently, magnetic resonance imaging (MRI) is the most widely used imaging modality for identifying brain tumor and other discrepancies. From these MRI images, we are able to determine the detailed anatomical information to examine the development of the human brain and to diagnose the various diseases. The tumor detection becomes most complicated for the huge image database. So a software approach is required to help the accurate and faster clinical diagnosis. The proposed system identifies and segments the tumor portions of the image successfully using MATLAB. The aim is to introduce an algorithm employed for performing useful operations on the MRI images such as filtering, pre-processing, morphological operation, K-means clustering, feature Extraction, Decision making system, SVM Classifier & Naïve Bayes. This work helps to save the time of both pathologist and doctor. The experimental results show that the proposed method detects tumor areas efficiently in the image of the brain.