ESTRO 2023 - Abstract Book

S1634

Digital Posters

ESTRO 2023

3 mm), a mean segmentation DSC of 0.78, a mean COB error of 2.7 mm, and a mean box DSC of 0.83, with an inference time of 300 ms/frame. Patients with tumor motion amplitude above 5.9 mm (median amplitude of test set) showed slightly better results in all four metrics than those with lower tumor motion amplitude.

Conclusion We implemented a patient-specific refinement approach using a pre-trained baseline model, which is suitable for real-time markerless lung tumor tracking. The implemented approach also provides quantitative criteria to select patients eligible for markerless tracking based on the PCT.

PO-1890 Dosimetric impact of setup and geometric uncertainties for free-breathing lung photon radiotherapy

N. Olofsson 1 , K. Wikström 2 , A. Flejmer 3 , A. Ahnesjö 1 , A. Dasu 4

1 Uppsala University, Immunology, genetics and pathology, Uppsala, Sweden; 2 Uppsala University Hospital, Medical physics, Uppsala, Sweden; 3 Uppsala University Hospital, Radiation oncology, Uppsala, Sweden; 4 The Skandion Clinic, Medical physics, Uppsala, Sweden Purpose or Objective Intrafractional respiratory motion is a concern for lung tumor radiotherapy but full evaluation of its impact is hampered by the lack of images representing the motion over the full treatment time. This study presents a novel evaluation using images acquired over realistic treatment times to investigate the dosimetric impact of respiratory motion during free breathing in photon radiotherapy. The image sequences were acquired at three occasions and were combined with simulated patient shifts to create a large set of probability weighted treatment scenarios for evaluation of different treatment planning strategies. Materials and Methods Three planning methods were compared all aiming to deliver 54 Gy as median dose (D50%) in three fractions to the target: one denoted FLU with homogeneous fluence to the PTV, one denoted ISD with isodose prescription to the PTV combined with a high central dose, and one denoted IRN with concurrent renormalization. As basis for simulated dose tracking in free-breathing we used cine-CT images of 14 patients with small lung tumors. Images were acquired during eight minutes at each occasion and used together with simulated patient shifts to create sets of probability weighted combinatorial treatment scenarios for each patient and treatment planning strategy. Using 15 different patient shifts, including the nominal, allowing repetitions and disregarding the order it can be shown that the number of treatment scenario combinations per patient and planning methods is 16 215 for a treatment in three fractions.