Advanced search
Start date

Federated learning for validating machine learning models trained in different hospital networks

Grant number: 22/16683-9
Support Opportunities:Scholarships in Brazil - Post-Doctoral
Effective date (Start): March 01, 2023
Effective date (End): January 31, 2024
Field of knowledge:Physical Sciences and Mathematics - Computer Science
Acordo de Cooperação: European Commission (Horizon 2020)
Principal Investigator:Paulo Mazzoncini de Azevedo Marques
Grantee:Gilson Yuuji Shimizu
Host Institution: Faculdade de Medicina de Ribeirão Preto (FMRP). Universidade de São Paulo (USP). Ribeirão Preto , SP, Brazil
Associated research grant:21/06137-4 - Predicting cardiovascular events using machine learning, AP.R


While several machine learning (ML) models that use electronic health record (EHR) data have been developed in recent years, there is a great need for external validation of these models. Prediction models may perform well in a center with a population similar to the training data, but may perform worse in centers with different patient characteristics. To develop generalizable models and discriminate equally well between different cohorts, multicenter studies should be considered. However, these multicenter studies are often limited if they require sharing patient data in a centralized location. Even if data is anonymized before being transferred, there is always some risk of anonymity being compromised on certain types of data. Approaches based on federated learning circumvent these limitations by sharing models and metrics instead of data, making it possible to improve the generalizability of prediction models while safeguarding patient privacy. This project proposes implementing a federated learning model and its evaluation as a tool to optimize the generalization capacity of machine learning models in different hospital networks. (AU)

News published in Agência FAPESP Newsletter about the scholarship:
Articles published in other media outlets (0 total):
More itemsLess items

Please report errors in scientific publications list using this form.