ESTRO 2025 - Abstract Book

S2889

Physics - Dose prediction, optimisation and applications of photon and electron planning

ESTRO 2025

3684

Mini-Oral Evaluation of different automation techniques in clinical treatment planning Francesca Nella, Lotte Wilke, Hubert Gabrys, Madalyne Day, Riccardo Dal Bello, Nicolaus Andratschke, Stephanie Tanadini-Lang Department of Radiation Oncology, University Hospital Zurich, Zurich, Switzerland Purpose/Objective: Our department has implemented several automation techniques for radiotherapy treatment planning. To further enhance and streamline the use of these tools, we conducted a perspective analysis of their current utilization. Material/Methods: Over a three-month period, treatment planners completed a brief questionnaire after finalizing each treatment plan. The questionnaire asked which automation tools, if any, were used in the planning process, and required reasons if no automation was utilized; the number of optimizations was also recorded. The planners were free in the choice of tool used. A script-based tool was available for plan preparation (including automatic treatment couch setup, creation of auxiliary structures, and initial plan setup with optimization objectives). For plan optimization, the following tools were available: predefined optimization templates (OT), a fluence-based forward optimization tool (EZ Fluence, EZF), a library-based optimization tool (RapidPlan, RP), and a specific optimization tool for brain metastases (HyperArc, HA). The script captured this information along with the treatment site and planner, allowing for evaluation across different treatment types and planners. Results: During the three-month period, 456 questionnaires were completed out of a total of 583 plans created (78%). In 369 cases (81%) some kind of automation was used. The plan-preparation script was used in 153 of the 456 cases (34%). Automation tools for plan optimization were applied in 288 (63%) plans, specifically: an optimization template in 111 cases, EZ-Fluence in 32 cases, RapidPlan in 137 cases and HyperArc in 8 cases. A median of 2 optimizations was used for modulated treatment-plans. Figure 1 presents the use of automation across different treatment techniques, while Figure 2 shows usage by different treatment planners. Main reasons for not using the plan preparation script were failure of the script due to various reasons, that the body-contour had to be unapproved prior to usage and usage of non-standard optimization-structures. Main reasons for not using automation during the optimization were complex re-irradiations and very large target volumes (>40cm) or cases where no automation tool was available.