Classification of the Type of Brain Tumor in MRI Using Xception Model

Abstract

A brain tumor is recognized as one of the most invasive things to operate on. Cancer develops inside the brain due to unregulated and aberrant cell partitioning. The recent breakthroughs in deep learning greatly aided the medical imaging sector in diagnosing numerous diseases. In MR images, visual learning and image recognition have been used to classify the type of brain tumor. The researchers utilized a Convolutional Neural Network (CNN) approach, Data Augmentation, and Image Processing to organize brain MRI scans as cancerous or non-cancerous. Using the transfer learning method, the researchers compared the performance of the primary CNN model to that of pre-trained CNN and Xception models. However, the experiment was conducted on a limited dataset. Later results reveal that the model’s accuracy result is very effective and has a meager complexity rate, attaining 96% accuracy on the Xception Model.