Automatic Detection and Classification of Enhanced Brain Tumor Using Machine Learning Algorithm

The early detection and proper treatment of brain tumor are essential to prevent permanent damage of brain. Present study proposes an automatic and effective approach to detect brain lesion in early stage that refers to the process of automated contrast enhancement of magnetic resonance (MR) brain images by incorporating simple power law transformation followed by segmentation and identification of the region of interest (ROI) using fuzzy c-means clustering technique and then finally classification of ROI into benignancy/malignancy classes by capturing six significant morphological feature selection. Finally, benignancy/malignancy of masses is examined and assessed by using well-known receiver operating characteristic method of ANN classifier based on significant feature selection. The result of the proposed method is enterprising with very low computational time and accuracy of 90.8%.