In association with KQCodes, ARC proudly presents the TechSocial Series. In October, we will be joined by Alexandre Triay Bagur & Rafael Dias, Senior AI Engineers at King’s College London & London AI Centre to talk about - FLIP: An open-source federated learning interoperability platform for collaborative AI in healthcare.
A series of FREE, informal TALKS, DISCUSSION and PIZZA! Open to anyone interested in computational research methods, technology and innovation, this series covers a broad range of tools, programs, digital environments & language.
Join us every month at 90 High Holborn
Abstract
Hospitals hold rich, multi-modal data with enormous potential for AI research, yet privacy regulations, institutional policies and technical barriers keep it largely inaccessible — and models trained on curated research cohorts often fail to generalise to real-world clinical populations. FLIP (Federated Learning Interoperability Platform), developed at the London AI Centre for Value-Based Healthcare, addresses this by enabling AI models to be trained across multiple NHS Trusts without patient data ever leaving each Trust's secure environment. At each site, data pipelines harmonise imaging and electronic health record data from over 20 NHS source systems into the OMOP common data model; researchers run federated cohort queries from a central hub, curate imaging data in XNAT, and train and evaluate models using federated learning frameworks such as NVIDIA FLARE and Flower. In this talk we will walk through FLIP's architecture and how it has evolved, share practical lessons from deploying research infrastructure inside NHS Trusts, and discuss our recent open-sourcing of the platform and how the community can build on it.
About the speaker
Alexandre Triay Bagur - Senior AI Engineer - King’s College London & London AI Centre
Prior to joining the Department as a Senior AI Engineer, Alex graduated with an MSc Biomedical Engineering from Imperial College London, focusing on Medical Physics and Imaging, and a DPhil from the University of Oxford. He had also worked as a Senior Imaging Scientist at Perspectum, developing Magnetic Resonance Imaging (MRI) methods for the quantification of liver disease, producing multiple journal articles and patents. Alex is a member of the London AI Centre for Value-Based Healthcare (aicentre.co.uk) and a core developer of FLIP, the Federated Learning Interoperability Platform that enables AI research using multi-modal NHS data.
Rafael Dias - Senior AI Engineer on Foundational Models for Healthcare - King’s College London & London AI Centre
Dr. Rafael Dias is a Senior AI Engineer on Foundational Models for Healthcare, with a strong background in Machine Learning, including both industry and academic experience. He holds a PhD in Machine Learning applied to Astrophysics and has over a decade of professional experience in Python programming. His scientific contributions include authoring ML book chapters and articles in Q1 journals. Dr. Dias is pioneering the development and deployment of advanced AI tools in collaboration with NHS Trusts, universities, and industry partners to enhance clinical decision-making and improve patient care pathways. He has been instrumental in the success of the Federated Learning Interoperability Platform (FLIP), a multi-NHS Trust data connectivity platform designed for scalable and privacy-preserving AI development. A core focus of his work involves leading the development of novel vision-language foundational models by integrating medical imaging (X-rays, CT, MRIs) with radiological reports. This work is enabling critical downstream applications such as automated report generation, image synthesis from reports, and discrepancy identification between reports and scans. He also contributes to prominent open-source projects like MONAI Core.
Visit our TechSocials site to find more information regarding future dates, times and registration.
Make sure to subscribe to the ARC mailing list if you want to receive updates for future events.
90 High Holborn
WC1V 6LJ
London
Begin:
End:
Add to Calendar