Detection of Brain Tumor in Medical Images Using Graph Neural Networks

• Grant number: 24/06012-5
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: June 01, 2024
• End date: May 31, 2025
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Zhao Liang
• Grantee: André Kenji Hidaka Matsumoto
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil

Abstract

Brain cancer ranks as one of the global causes of death, and its early detection is challenging due to the diversity of sizes and shapes present in Magnetic Resonance Imaging (MRI). Moreover, identifying the specific type of tumor is a complex process that demands expertise. The use of artificial intelligence and machine learning in computer-aided diagnosis systems becomes essential to enhance accuracy and reduce the workload of radiologists.Graph Neural Networks (GNNs) have sparked considerable interest in the field of Deep Learning, thanks to their unique ability to deal with unstructured data types and their pattern invariant recognition ability. In this project, the development of a system that converts images into graphs is proposed, implementing a GNN called TransGNN, with the aim of detecting brain cancer. It is expected that this technique will be able to identify brain tumors, regardless of their variety in size and shape.