ESTRO 2025 - Abstract Book

S4426

Late-breaking abstracts

ESTRO 2025

For GTV-T, there was a clear tendency for the spread of changes over the population to be smaller in the intervention arm than in the control arm, with fewer cases of large changes, for all metrics. For GTV-N, the intervention arm and control arm were more similar with a slight tendency for larger changes in the intervention arm for APL. Conclusion AI assistance was shown to be a useful tool for GTV contouring for head and neck cancer in clinical practice with performance at least as good as manual contouring. This result paves the way for including AI assistance in standard clinical routine in the future to reduce manual workload.

Keywords: Head and neck cancer, target contouring, AI