Applied Computing Events

Title: Real-World Federated Learning for Healthcare: Achievements so far & open research questions

Abstract: This talk will focus on Federated Learning (FL) for training, evaluating, and monitoring diagnostic, prognostic, and predictive Healthcare AI models. It will compare the current centralized learning with the proposed decentralized/FL paradigm of multi-institutional AI studies, in recognition of the need for large and importantly diverse data to capture diverse patient population demographics. It will further provide an overview of successful real world FL initiatives for big data analysis. Finally, it will present the current open questions that need to be addressed to facilitate real-world deployment of FL at scale across distinct geographic locations, while keeping in mind the health system environment.

Bio: Dr Spyridon Bakas is an Endowed Chair Associate Professor at Indiana University School of Medicine, the Director of the Computational Pathology Division, & the Director of the Center for Federated Learning in Medicine, with his primary appointment in the Dept of Pathology & Laboratory Medicine, as well as secondary appointments in the Depts of Radiology & Imaging Sciences, Biostatistics & Health Data Science, Neurological Surgery, & Computer Science. His group focuses on developing, applying, & benchmarking computational algorithms in medical imaging, towards improving disease assessment and diagnosis in the current clinical practice. He is leading an NIH-funded group focusing on image quantification, imaging-genomics, and federated learning (FL), towards enabling treatment selection models customized on an individual patient basis, while addressing health disparities and inequities. His group has introduced FL in healthcare in 2018 and driven the largest to-date real-world FL studies including data from 71 sites across 6 continents. He has co-authored >130 peer-reviewed manuscripts and >90 abstracts, with collaborators across academic ranks and disciplines. He is a member of the MICCAI Society Board of Directors, the President of the MICCAI Special Interest Group (SIG) on Computational Pathology, the Vice Chair for Benchmarking & Clinical Translation in the MLCommons’ Medical group, the co-lead of the AI-RANO working group, and has served as the organizer and chair of numerous challenges, workshops, and tutorials at both technical and clinical scientific meetings.

How to attend: Either turn up to the event on the day, or if you want to attend online then please contact Adam Shephard (adam.shephard@warwick.ac.uk) for more details.