New AEI-Funded Project to use Artificial Intelligence to study the impact of extreme weather on diabetes

projects
We are starting the fDICE project to combine Artificial Intelligence, large-scale health records and environmental data to investigate how extreme weather affects diabetes and cardiovascular diseases
Author

Jordi Fonollosa

Published

September 20, 2026

Modified

September 20, 2026

Climate change and the growing prevalence of type 2 diabetes mellitus (T2DM) represent major challenges for public health. Extreme temperatures, heatwaves, cold spells and air pollution can aggravate cardiometabolic conditions. However, these environmental factors are rarely incorporated into conventional health risk prediction models.

The fDICE project aims to address this challenge by combining Artificial Intelligence, large-scale electronic health records and environmental data to better understand and predict how extreme weather affects diabetes and cardiovascular health.

The project is just starting and will develop advanced AI models based on transformer architectures, learning from longitudinal patient records enriched with high-resolution environmental information. By integrating clinical histories, diagnoses, treatments, laboratory results and environmental exposures, these models will identify patterns of disease progression, predict cardiometabolic complications and explore how weather conditions influence individual health trajectories.

A main objective of the project is to develop more accurate and generalisable prediction tools to identify vulnerable populations, anticipate weather-related health complications and support more personalised healthcare.

The fDICE project (Foundation Models for DIAbetes Multimorbidity and ClimateEnvironmental Risk) has been funded by the Spanish State Research Agency (Agencia Estatal de Investigación, AEI) through its 2025 Knowledge Generation Projects call, reference PID2025-173811OB-I00. The project is coordinated by the Institut de Recerca i Innovació en Salut (IRIS), at the Universitat Politècnica de Catalunya (UPC), in collaboration with researchers from the Institut Català de la Salut and from Fundació Mútua de Terrassa.