At LUMA@UCL, we build AI models that learn from the human body, and our research focuses on structure, reasoning, and clinical collaboration.
We study how machines can understand anatomy, motion, and clinical thinking — so that medical AI becomes more useful, trustworthy, and safe in real-world healthcare.
We build generative models of cardiac anatomy and motion — to support simulation, personalisation, and digital twins.
We design collaborative agent systems that interpret and reason across clinical data, inspired by how care teams work.
We develop learning methods that incorporate anatomical structure into representation and model behaviour.
William Gibson
Luma Lab
Copyright © 2025 Luma Lab - All Rights Reserved.