ESTRO 2024 - Abstract Book

S3560

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

ESTRO 2024

Conclusion:

Clinical compliance was achieved with all automatically generated SBRT lung plans and the majority of SBRT pelvic LN treatment plans. A PTV threshold of 3.6 cm 3 can be used as quality criteria in clinical settings for pelvic LN treatment plans and 4.0 cm3 for lung plans. The results of our study demonstrated the promising performance of Halcyon for pelvic and lung SBRT, although plan-specific QA is required to verify machine performance during plan delivery.

Keywords: O-ring linac, SBRT, auto planning

References:

1. Reshko LB, Richardson MK, Spencer K, Kersh CR. Stereotactic Body Radiation Therapy (SBRT) in Pelvic Lymph Node Oligometastases. Cancer Invest. 2020;38(10):599-607.

2. Pokhrel D, Tackett T, Stephen J, et al. Prostate SBRT using O-Ring Halcyon Linac - Plan quality, delivery efficiency, and accuracy. J Appl Clin Med Phys. 2021;22(1):68-75

3. Pokhrel D, Webster A, Stephen J, St Clair W. SBRT treatment of abdominal and pelvic oligometastatic lymph nodes using ring-mounted Halcyon Linac. J Appl Clin Med Phys. 2021;22(6):162-171.

1377

Digital Poster

Dose Prediction for Prostate Radiotherapy Planning

Edouard Delasalles 1 , Rémi Vauclin 1 , Elie Mengin 1 , Baris Ungun 1 , Madalina-Liana Costea 2 , Norbert Bus 1 , Nikos Komodakis 3 , Pascal Fenoglietto 4 , Gorkem Gungor 5 , Nikos Paragios 6 1 TheraPanacea, Physics, Paris, France. 2 TheraPanacea, Clinical affairs, Paris, France. 3 University of Crete, Computer vision, Heraklion, Greece. 4 Institut du Cancer de Montpellier, Department of Radiation Oncology, Montpellier, France. 5 Acibadem MAA University School of Medicine, Department of Radiation Oncology, Istanbul, Turkey. 6 TheraPanacea, CEO, Paris, France

Purpose/Objective:

Radiotherapy (RT) planning is a long iterative process that requires time and effort from highly trained doctors, dosimetrists and physicists. With the advancement of treatment technology, the modularity of radiotherapy machines has increased dramatically. Techniques such as Volumetric Arc Therapy (VMAT) are incredibly powerful, but require even more time and skills to achieve good results. In order to reduce this time and bring more automation to the planning process, we present in this work a dose volume prediction neural network. It takes the form of a Convolutional Neural Networks (CNN) that uses the patient CT image with associated planning target volume (PTV) and organ-at-risk (OAR) contours to predict a dose