A Segmentation Based Automatic System for Brain Tumor Classification from CT Scan Images

Brain tumors are the most forceful and devastating variant of tumor and hence, its correct recognition at an early stage took after by treatment is its only cure. Brain tumors has a variation what's more, complex structure and thus its classification is difficult. This paper disclosed a framework for effective brain tumor medical image classification. Here, mean to utilize a huge datasets to design classifier. A Segmentation based automatic system for brain tumor classification from CT scan comprises of three phases, in particular, Preprocessing, Segmentation, Feature extraction, and Classification. Five distinctive classifiers, Naive Bayes, K-Nearest Neighbors(KNN), Multilayer perceptron(MLP), Support Vector Machine(SVM) and J48 classifiers are used for testing the discriminating power of the feature set. MLP classifier achieve a cross acceptance precision of 85.2041% over all other classifiers.