Brain tumor grade detection by using ANN

Abstract

Brain tumor is death threatening disease, the early detection helps to extend the life time. Practically, the micro analysis of tissues only confirms the disease severity. Hence, a computer aided method is proposed to categorize brain tumors as benign and malignant by using Feed forward network. The existing works assign features as much of pixel counts. However, the proposed method eliminates the feature size by using significant features according to the tumor and brain anatomy. The network attains 96% of accuracy and compared against existing works.