Detection of Brain Tumor using CNN
A CNN model is proposed and evaluated against ROC Curve, Confusion Matrix, and Accuracy Curve and it is observed that the proposed model achieved better accuracy when compared with state-of-the-art models.
Abstract
Brain, which is mainly composed of billions of neurons, hosts the most complicated processes that occur in the body. A brain tumor is a tissue growth in the brain that grows in an area of the brain, which should not be significantly increased. To analyze brain tumors, an open source dataset available in Kaggle is used in the paper. There are photos in the dataset and all images are in different colors and dimensions. All noisy images are converted into gray scale using GaussianBlur filter with dimensions 256*256. Then the dataset is divided into 80% for training and 20% for testing. A CNN model is proposed and evaluated against ROC Curve, Confusion Matrix, and Accuracy Curve. It is observed that the proposed model achieved better accuracy when compared with state-of-the-art models.