A Deep Learning Based Dual Model Developed for Automated Detection of Glioma-brain Tumor

Abstract

Advances in machine learning (ML) and artificial intelligence (AI) are truly groundbreaking. We propose a fully automatic way to classify whether individuals contain lower-grade gliomas (brain tumors) using a deep-learning segmentation algorithm in this research paper. There are two deep-learning models in our paper. The first model detects whether the brain contains a tumor. The second model generalizes the tumor's location. In recent years, deep learning for automated brain segmentation has progressed to the level of success of a professional radiologist. Because of the class imbalance, we used the Tversky loss function. Experimental results show a better performance of the proposed model compared to the existing deep-learning models.