Federated Learning Could Have “A Catalytic Impact Towards Precision Medicine”

Researchers from Washington University School of Medicine in St. Louis are exploring how to facilitate multi-institutional collaborations without sharing private patient data.

Daniel S. Marcus, PhD (above), professor of radiology, and Mikhail V. Milchenko, PhD, instructor in radiology — both members of the Computational Imaging Research Center, part of the School of Medicine’s Mallinckrodt Institute of Radiology — published a study in Nature featuring the potential of federated learning and its “catalytic impact towards precision medicine.”

Federated learning is a novel paradigm that leverages all available data without sharing data between institutions. They showed 99% of the model quality achieved with centralized data.

Read the paper here.