WP A2 develops AI-driven MRI-guided adaptive radiotherapy. MRI-guided radiotherapy makes it possible to see soft tissues during treatment, which is especially valuable when tumours or organs move or change shape. However, today’s MRI-guided treatment sessions can still be time-consuming because many steps require expert manual input.

This work package aims to make MRI-guided treatment faster, more automated and more personalised. It combines AI-assisted tumour delineation, automated treatment planning, real-time tracking, motion management, audiovisual biofeedback and dose reconstruction. The goal is to support treatment decisions based on what is actually visible during treatment and on the dose that is actually delivered.

The work has two main directions. The first is real-time anatomy-guided radiotherapy, where treatment adapts to anatomy and motion. The second is dose-guided radiotherapy, where information about the delivered dose is used to guide later treatment decisions. Together, these developments could help reduce margins, shorten treatment sessions and support more personalised fractionation strategies.