Name in original language
-
Initiative overview
The AI-driven proton therapy planning project, developed through collaboration between University Medical Center Groningen (UMCG) and RaySearch, aims to automate the creation of robust treatment plans for head and neck cancer patients. The system uses a machine learning model trained on historical data from 80 patients to infer dose distributions and generate intensity-modulated proton therapy (IMPT) plans. The primary goals of the project are to reduce manual workload, improve planning efficiency, and maintain clinical quality and robustness. The framework supports rapid plan generation—under 20 minutes—and is designed to facilitate adaptive therapy and modality comparisons between photon and proton treatments.

























