Artificial intelligence in medical image processing: Diseases classification

Abstract

This research focuses on developing a robust model to classify tumors into categories (glioma, meningioma, pituitary, no tumor) accurately and efficiently. The work dives in the history of medical image classification and its evolution. It also involves leveraging deep learning techniques, to assist in the detection and categorization of brain tumors from MRI scans. A CNN model was implemented in this work, analyzed and well described. A transfer learning model was also used. From the results of each model, they will be compared. A discussion was initiated on the limitations in this field, and some considerations were brought up.