First, Do No Harm — In Silico: How Digital Twins Are Rewiring Medical Evidence, Patient Safety, and the Regulatory Mind

Speaker: Professor Alejandro F. Frangi
Bicentenary Turing Chair of Computational Medicine
University of Manchester

Title: First, Do No Harm — In Silico: How Digital Twins Are Rewiring Medical Evidence, Patient Safety, and the Regulatory Mind

Date: Thursday, 14 May 2026

Time: 11:00am - 12:30pm

Venue: Room 4472 (via lift 25/26), HKUST

Abstract:

Randomised controlled trials are medicine's gold standard, yet they remain slow, narrow, and blind to the full diversity of patients. In silico trials — virtual experiments on digital twins of populations, devices, and clinical workflows — promise a faster, safer, and more inclusive evidence pipeline, but only if we can trust them. This talk walks the boundary between promise and breakage. Using concrete exemplars in neurovascular devices, orthopaedic implants, and cardiovascular interventions, I trace where in silico methods already replicate and extend physical trials, where they fracture under data drift and model misspecification, and why the hardest problem is no longer simulation fidelity but regulatory confidence. I introduce the concept of a "Regulatory Airlock" — a pre-competitive, cross-sector framework for co-producing the standards that turn computational models into admissible evidence. The talk closes with a view of the emerging technical backbone: generative AI for virtual populations, multi-scale physiological coupling, and the governance structures required for a self-improving, trustworthy in silico ecosystem.

Biography:

Professor Alejandro F. Frangi holds the Bicentenary Turing Chair of Computational Medicine at the University of Manchester, where he directs the Centre for Computational Imaging and Modelling in Medicine and the UK Centre of Excellence for In Silico Regulatory Science and Innovation (UK-CEiRSI). As a Royal Academy of Engineering Chair in Emerging Technologies, he pioneered at-scale in silico trials that have replicated and expanded landmark clinical studies in neurovascular and cardiovascular disease. His work bridges mechanistic modelling, generative AI, and regulatory science to build trustworthy digital twins for medical evidence. He is a Fellow of the Royal Academy of Engineering and advises regulators and industry on computational evidence frameworks.