AUTOPT: Fully-automated multi-criterial planning for intensity-modulated proton therapy (IMPT) for clinical practice


Principle Investigators: Prof. dr. Coen Rasch (LUMC), Prof. dr. Ben Heijmen (Erasmus MC)

PhD students: Merle Huiskes (LUMC),  W. Kong (Erasmus MC)

Funding: HollandPTC-Varian 2019


Treatment plan generation for intensity-modulated proton therapy (IMPT) is highly complex, especially if it includes selection of favourable beam geometries. Currently, treatment plans are generated in an iterative trial-and-error procedure (‘manual planning’). Each iteration is time-consuming, which severely restricts the total number of iterations that can be performed for an individual patient, risking sub-optimal quality of the final plan. This quality is also dependent on the experience and subjective choices made by the manual planners. To mitigate as much as possibly issues with plan quality, the current manual planning in proton therapy centres is unavoidably associated with a high workload.

This project develops and investigates fully-automated, multi-criterial and robust planning for IMPT for clinical use. As a starting point and core application for the investigations we will use the iCycle optimizer which has been extensively studied and validated for automated treatment planning (autoplanning) for photon beams. Options for integrated computerized beam angle selection for IMPT, taking into account robustness requirements, will be added to iCycle.