ESTRO 2025 - Abstract Book

S2893

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

ESTRO 2025

processing for multiple patients with a single mouse click. The FT process efficiently produced clinically acceptable plans, laying the foundation for broader applications. A script that automates treatment planning for multiple patients simultaneously was developed and evaluated. This approach optimizes computational resources by scheduling plan calculations during low system load periods, such as nighttime, paving the way for more efficient and scalable workflows. Material/Methods: An executable script based on ESAPI, compatible with Varian Eclipse TPS v15.6, was developed to fully automate treatment planning for multiple prostate cancer patients. This script consolidates multiple functionalities: invoking a knowledge-based planning (KBP) model, applying a FT process, and enabling batch processing for multiple patients simultaneously. The script was tested on a cohort of 25 patients, randomly selected from a clinical database. Each plan was evaluated against standard clinical dose constraints to determine plan acceptability. Additionally, the monitor units per gray and de complexity metric (CM) [2] were used to compare plans. Results: Among the 25 prostate cases tested using this script, all constraints of 6 treatment plans were met with only the KBP model. The FT process addressed the unmet constraints in another 14 cases, resulting in an increase in the number of total plans meeting all constraints from 6/25 to 20/25. The mean MU/Gy was 364±42 for KBP and 349±45 for FT, while the mean CM was 0.18±0.05 for KBP and 0.17±0.05 for FT. Figure 1a and 1b respectively display the comparison between the result of the KBP (first loop) and FT (last loop) regarding the MU/Gy and the CM for the 14 cases that required the FT process to generate plans meeting all constraints.

Conclusion: The single-click batch auto-planning script offers a promising solution for automating prostate planning. The script effectively improved the number of plans meeting clinical constraints while demonstrating a reduction in MU/Gy without increasing plan complexity. These advancements allow for more efficient dose delivery without compromising dosimetric accuracy. By streamlining clinical workflows, reducing planning time, optimizing computer