Brain Tumor Segmentation Using Machine Learning Classifier

Using brain magnetic resonance Images a machine based software is developed for segmentation and to classify tumor type as benign and malignant. In this study, the MRI image enhanced using contrast improvement technique, double thresholding is done using morphological operations and skull striping process is used mainly to remove unwanted non cerebral tissues from MR image, The brain tumor segmentation is done in a slice of Magnetic Resonance (MR) Image where massive abnormal cells are localized and tumor region that are sliced are segmented by machine learning classifiers like KNN. fuzzy C-mean. k-means. Feature are derived using GLCM and those features are trained in such a way it produce accurate segmentation of tumor region is done in less computation time and therefore, in proposed system, features derived from the GLCM. so that the segmentation of tumor region using triple technique K-means. KNN and FCM can be done accurately and efficiently. Accuracy and error rate is calculated for brain MRI image using triple technique Means, FCM and KNN.