Brain Tumor Semantic Segmentation and Classification using Deep Learning techniques

Abstract

ABSTRACT : Brain cancer, speciőcally Glioma, is a devastating disease with a very low chance of survival. In fact, only 3.6% of patients diagnosed with high-grade Glioma survive beyond őve years. For medical professionals, accurately identifying and categorizing brain tumors into different classes is vital when it comes to diagnosing and planning the appropriate treatment for patients. Magnetic resonance imaging (MRI) is commonly used to examine brain tumors in clinical practice. Fortunately, deep learning methods have shown remarkable potential in effectively segmenting brain tumors and have yielded promising results in various biomedical applications. This study examines brain tumors semantic segmentation and classiőcation that used deep learning algorithms in medical technology applications such as Unet, Resnet, VGG net. We initiate by providing an overview of the basic principles of deep learning. Subsequently, we delve into the applications of deep learning in the őeld of diagnosing and segmenting brain tumors using magnetic resonance (MR) images. Lastly, we conduct a comparative analysis of various approaches that have been explored, highlighting their respective őndings and outcomes.