THEORETICAL FRAMEWORK FOR THE DEVELOPMENT OF ARTIFICIAL NEURAL NETWORKS AIMED AT ASSISTING IN THE DIAGNOSIS OF HODGKIN'S LYMPHOMA

Abstract

This study focuses on a theoretical framework for the development of artificial neural networks (ANNs) with the aim of assisting in the diagnosis of the hematological disease Hodgkin's Lymphoma (HL). Initially addressing the architectures and algorithms of neural networks, as well as learning and training methods. The study elucidates how these techniques can be applied to identify complex patterns amidst human DNA genetic mutations, showcasing tools and concepts to achieve the objective. Concurrently, the article explores Hodgkin's Lymphoma, examining the involved genes and gathering information provided through public databases. The integration of this genetic information with ANNs has the potential to create a robust and efficient tool. By establishing this multidisciplinary theoretical base, the article lays the foundations for future developments of systems that seek to improve the diagnosis of Hodgkin's Lymphoma and other predominantly genetic diseases, which can, consequently, achieve more effective and personalized treatments.