ESTRO 2024 - Abstract Book

S3565

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

ESTRO 2024

[1] M. Hussein, B. J. M. Heijmen, D. Verellen, and A. Nisbet, “Automation in intensity modulated radiotherapy treatment planning—a review of recent innovations,” The British Journal of Radiology, vol. 91, no. 1092, p. 20180270, Dec. 2018, doi: 10.1259/bjr.20180270. [2] P. Meyer et al., “Automation in radiotherapy treatment planning: Examples of use in clinical practice and future trends for a complete automated workflow,” Cancer/Radiothérapie, vol. 25, no. 6–7, pp. 617–622, Oct. 2021, doi: 10.1016/j.canrad.2021.06.006. [3] M. Zelefsky, “The GETUG 70 Gy vs. 80 Gy randomized trial for localized prostate cancer: feasibility and acute toxicity,” Urologic Oncology: Seminars and Original Investigations, vol. 23, no. 4, p. 307, Jul. 2005, doi: 10.1016/j.urolonc.2005.05.014.

1411

Poster Discussion

Quantitative and qualitative evaluation of an automated planning solution for prostate radiotherapy

Madalina-Liana Costea 1 , Baris Ungun 2 , Rémi Vauclin 2 , Edouard Delasalles 2 , Elie Mengin 2 , Norbert Bus 2 , Gorkem Gungor 3 , Matthias Moll 4 , Salvatore Cozzi 5 , Vincent Gregoire 5 , Pascal Fenoglietto 6 , Nikos Paragios 7 1 TheraPanacea, Clinical Affairs, Paris, France. 2 TheraPanacea, Physics, Paris, France. 3 Acibadem MAA University School of Medicine, Department of Radiation Oncology, Istanbul, Turkey. 4 Medical University of Vienna, Department of Radiaton Oncology, Vienna, Austria. 5 Centre Léon Berard, Department of Radiation Oncology, Lyon, France. 6 Institut du Cancer de Montpellier, Department of Radiation Oncology, Montpellier, France. 7 TheraPanacea, CEO, Paris, France

Purpose/Objective:

Volumetric modulated arc therapy (VMAT) is widely used today for radiotherapy (RT) treatment planning. However, manual treatment planning (MP) task is labor exhaustive and highly dependent on planner’s skills and experience. With the emergence of knowledge-based treatment planning approaches, deep learning dose prediction and dose mimicking solutions are highly sought for reducing the planning time without compromising plan quality. We propose a fully-automated treatment planning (AP) approach for prostate cancer treatments that requires no expert intervention between contour approval and dosimetric review for plan sign-off.

Material/Methods:

We have developed a treatment planning pipeline that takes as input a planning CT with organs-at-risk (OAR) and planning target volume (PTV) contours, the targeted linac machine and the prescription dose. The primary components are (i) dose prediction by a deep learning model trained on 123 clinical cases and (ii) direct aperture VMAT plan optimization that seeks to mimic the predicted dose. An end-to-end clinical evaluation study was performed on another 25 cases. The RT plans generated by the pipeline were calculated using a Collapsed Cone Convolution engine and the obtained RTdoses were compared with the reference doses from MP.

First, a quantitative evaluation was performed based on dose-volume histogram (DVH) points and plan parameters metrics (monitor units (MU) and modulation complexity score (MCS)). Paired Wilcoxon signed-rank