Federated Learning for Survivability Analysis in Personalized Healthcare

16 Maggio 2025
Share on facebook Share on twitter
27
VIEW

Meeting Room Dipartimento di Ingegneria dell'Informazione
Università di Pisa
Via Caruso 16 - ground floor

Sajal K. Das, Fellow of IEEE
Curators’ Distinguished Professor and Daniel St. Clair Endowed Chair
Department of Computer Science
Missouri University of Science and Technology, USA

 

Abstract:
Enhancing healthcare outcomes in a privacy-preserving landscape demands innovative solutions that can integrate heterogeneous patient data from diverse sources. This talk will explore a novel Federated Learning (FL) framework for survivability analysis of patients, particularly cancer patients, leading to personalized well-being. The framework leverages geographic public health features, data heterogeneity, and trust mechanisms across healthcare institutions, as well as Cox Proportional Hazards (CoxPH) model and dynamic clustering. By enabling individual institutions to locally train on sensitive patient data — including location-based and public health statistics — our FL framework collaboratively learns a robust global survival model by protecting patient privacy and demonstrating strong performance metrics, such as concordance index (C-index), across synthetic and real-world datasets for cancer patients. The talk will emphasize techniques for efficient model aggregation, edge-centric optimization, handling heterogeneous feature spaces, and adapting recent breakthroughs in federated and semi-decentralized architectures. The end goal is to help build trust in collaborative healthcare ecosystems and reduce bias against smaller institutions.

 

Biography:
Sajal K. Das is a Curators’ Distinguished Professor of Computer Science and Daniel St. Clair Endowed Chair at Missouri University of Science and Technology, Rolla where he was the Chair of Computer Science Department during 2013-2017. Previously, he served as a Program Director at the NSF. His interdisciplinary research interests include CPS, IoT, drones, cybersecurity, machine learning, wireless and sensor networks, mobile and pervasive computing, smart environments (including smart health), edge/cloud computing, and applied graph theory and game theory. He has made fundamental contributions to these areas and published extensively in high quality journals and peer-reviewed conference proceedings, numerous book chapters, 4 books, and 5 US patents. Dr. Das has directed over $25 million in funded projects. According to Google Scholar, his h-index is 102 with more than 44,500 citations. He is the founding Editor-in-Chief of Elsevier’s Pervasive and Mobile Computing journal and Associate Editor of the IEEE Transactions on Dependable and Secure Computing, IEEE Transactions on Sustainable Computing, IEEE/ACM Transactions on Networking, and ACM Transactions on Sensor Networks. A founder of IEEE PerCom, IEEE WoWMoM, IEEE SMARTCOMP, and ACM ICDCN conferences, he has served as General and Technical Program Chair of numerous conferences. He is a recipient of 14 Best Paper Awards at prestigious conferences like ACM MobiCom and IEEE PerCom; and awards of excellence for teaching, mentoring, research and innovation including IEEE Computer Society’s Technical Achievement award and University of Missouri System President’s Award for Sustained Career Excellence. He has graduated 11 postdocs, 51 PhD and 31 MS thesis students. A Distinguished Alumnus of the Indian Institute of Science, Bangalore, Dr. Das is a Fellow of the IEEE, National Academy of Inventors (NAI), and Asia-Pacific Artificial Intelligence Association (AAIA).