Comparison of machine learning models and statistical methods for predicting survival in cancer patients

• Grant number: 25/03353-9
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: May 01, 2025
• End date: April 30, 2026
• Field of knowledge: Health Sciences - Collective Health - Epidemiology
• Principal Investigator: Tatiana Natasha Toporcov
• Grantee: João Vitor Morimoto Sesma
• Host Institution: Faculdade de Saúde Pública (FSP). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Associated research grant: 21/11794-4 - Cancer Control in the state of São Paulo (ConeCta-SP): from knowledge to action, AP.CCD

Abstract

In Brazil, there is a lack of studies that assess the validity of using Artificial Intelligence (AI) algorithms to predict the survival of cancer patients using large official databases. The Hospital Cancer Registry of the state of São Paulo (RHC/SP), based at the Oncocentro Foundation of the State of São Paulo (FOSP), represents a unique opportunity to conduct studies using this approach in the state of São Paulo, as it contains a historical series of more than 20 years of cancer cases in hospitals in São Paulo. The IC fellow will work on comparing the accuracy of AI algorithms and other statistical models, such as the Cox model, used to predict the survival of colorectal cancer patients in the state of São Paulo between 2000 and 2024. Their activities will include assessing the consistency of databases and data analysis; organizing a literature review on the topic, with the purpose of better interpreting the results from an epidemiological perspective; and producing scientific articles. (AU)