The convergence of digital twins, bio-mechanistic modelling, and artificial intelligence (AI) is revolutionizing the field of complex physiological systems, from cardiovascular to cancerous diseases. These systems allow for continuous monitoring and predictive modelling of patient health, helping clinicians explore treatment outcomes and interventions in a virtual setting before application in the real world. However, physiological systems are inherently complex, involving a multitude of interacting processes across various physical domains. The intricate interplay of these processes, occurring at different scales from molecular to organ-level, poses significant challenges for accurate and holistic modelling. Thus, a comprehensive framework that integrates multi-physics models and efficiently captures the underlying dynamics is required. The development of model reduction techniques to facilitate real-time simulations, alongside systematic approaches to explore subsystem behaviour, is critical to advancing digital twin applications in clinical settings.

This mini-symposium will address these challenges with a particular emphasis on computational oncology, where the integration of biological processes across multiple scales is key to understanding tumour growth and treatment response. The ability to model interactions from the genomic level to cellular metabolism and tissue microstructure is essential for accurate representations of tumour pathophysiology. As digital twins, bio-mechanistic models, and AI applications in oncology become more sophisticated, there is a growing need for models that not only simulate these complex systems but also do so in a manner that is reproducible, explainable, and shareable across research and clinical communities. Moreover, ensuring that these models are rigorous, interpretable, and scientifically valid remains a key challenge. This mini-symposium aims to bring together experts in the fields of physics, mathematics, and computational science to discuss and develop the necessary tools to advance digital twin approaches in oncology and beyond. By fostering collaboration between these disciplines, we aim to address the pressing need for more effective, explainable, and scalable digital twin solutions in complex physiological systems. The discussions of this session will contribute to the broader application of AI-driven modelling in a wide range of physiological systems, offering new pathways for personalized healthcare.