S1733
Physics - Dose prediction/calculation, optimisation and applications for particle therapy planning
ESTRO 2026
the automatically generated (AG)and manual plans were compared with the Wilcoxon signed-rank test. Plan quality was blindly scored on a four-point scale by two radiation oncologists, technologists, and physicists. For plan comparison, predicted and AG PT doses were evaluated against photon plans, as were the manual plan doses. PT was considered favorable if it achieved ≥ 5% dose reduction in the supratentorial brain or hippocampi. Results: Compared to manual PT plans, AG plans achieved improved CTV coverage (Table 1) with similar conformity indices (manual 0.75; AG 0.73). Several OOIs showed statistically significant but mostly clinically minor dose differences (median ≤ 0.25 GyRBE; Table 1). Eye structures showed larger variations, occasionally requiring manual adjustment. Depending on the observer, 60–94% of AG plans were rated acceptable or needing minor changes, versus 76–100% of manual plans (Figure 1A). Plan preference varied, but 60–70% of AG plans were considered at least comparable to the manual plan (Figure 1B). For photon versus PT comparison, both AG and predicted doses reproduced manual plan conclusions in 88% of cases.
Conclusion: The deep learning model accurately predicted PT– photon plan comparison outcomes. The dose- mimicking framework produced acceptable plans in most cases, supporting clinical implementation. Although DVH differences from manual plans were generally small, doses to eye structures may require manual refinement. References: [1] Eriksson O, Zhang T. Robust automated radiation therapy treatment planning using scenario - specific dose prediction and robust dose mimicking. Medical Physics 2022;49:3564–73. https://doi.org/10.1002/mp.15622. Keywords: deep learning; dose mimicking; neurological cancer Digital Poster Highlight 2472 Robust Prediction of Hematologic Adverse Event in Extended-Field Proton Therapy for Cervical Cancer:Insights from Real-World Daily Treatment Scenarios Wiwatchai Sittiwong 1,2 , Vossco Nguyen 2 , Molly Munro 2 , Deepali Purohit 2 , Courtney Reynolds 2 , Asma Sarwar 2 1 Division of Radiation Oncology, Department of Radiology, Faculty of Medicine Siriraj hospital, Bangkok, Thailand. 2 Radiotherapy, University College London Hospital, London, United Kingdom
Made with FlippingBook - Share PDF online