Detection of Brain Tumors from MRI using Gaussian RBF kernel based Support Vector Machine

Abstract

The Support Vector Machine (SVM) is a powerful classification technique that has been used extensively in the field of medical imaging. A model based on SVM with Gaussian RBF kernel is proposed here for the automatic detection of brain tumor from MRI images. Various textural characteristics of the MRI images of human brain are extracted to construct a feature set. These features sets are then used to train the classifier. The results obtained are compared with another powerful efficient classifier AdaBoost. AdaBoost classifies data according to the law of majority vote by base classifiers (ensemble learning). The comparative results show that though the difference between the performance measures is marginal, SVM gives higher precision and low error rates.