AI and Digital Twins in Healthcare: Synergies of Physics-informed Models and Machine Learning for Precision Medicine
Julia Camps. University of Oxford
Abstract
This tutorial explores the powerful intersection of artificial intelligence (AI) and digital twins, focusing on revolutionising healthcare applications. Participants will examine the key differences between biological digital twins and traditional industrial models, addressing the unique challenges of healthcare, such as patient variability, complex biological processes, and mechanistic uncertainty. The session will compare physics-informed machine learning with digital twins built from first principles using modelling and simulation, showcasing their respective strengths, limitations, and complementary roles. Through case studies in precision cardiology, we will illustrate how AI-enhanced digital twins improve diagnostic accuracy, predictive capabilities, and personalised treatment strategies. Additionally, participants will gain insights into overcoming critical challenges, including real-time data integration, computational costs, and ensuring high-performance outcomes in clinical settings.
Learning outcomes:
- Understand the fundamental differences between biological and industrial digital twins.
- Gain knowledge on the role of AI and physics-informed machine learning in developing medical digital twins.
- Know the pros and cons of traditional modelling and simulation versus statistical-based methods.
- Learn practical applications of digital twins in precision cardiology for diagnosis, prediction, and treatment.
- Explore strategies to address challenges like data integration, real-time performance, and model credibility.