RICCC presents federated ICU machine learning at SAIL 2026 in Puerto Rico

The RICCC team traveled to the Symposium on Artificial Intelligence for Learning Health Systems (SAIL) 2026, held May 5–8 in Río Grande, Puerto Rico. Data scientist Vaishvik Chaudhari and Dr. J.C. Rojas presented the lab's work on privacy-preserving, federated machine learning for the intensive care unit.

Led by Vaishvik Chaudhari, the poster — "Privacy-Preserving Multicenter Evaluation of ICU Clinical Decision Support: A Federated Framework Across Eight U.S. Health Systems" — demonstrates how ICU clinical decision support models can be evaluated across eight health systems without sharing patient-level data. Built on the Common Longitudinal ICU data Format (CLIF), the federated framework moves the model to the data at each site, enabling rigorous multicenter validation while keeping protected health information behind institutional firewalls.