S1794
Physics - Dose prediction/calculation, optimisation and applications for photon and electron planning
ESTRO 2026
MÁLAGA, Spain. 3 Servei de Física i Protecció Radiològica, Hospital Universitari Vall d'Hebron, Vall d'Hebron Barcelona Hospital Campus, BARCELONA, Spain Purpose/Objective: Prostate SBRT using an online adaptive CBCT-based procedure has been clinically described.1 We aim to assess the variations in target delineation performed by a single observer on the daily CBCT scans. Material/Methods: Twenty patients treated with online adaptive SBRT were randomly included in this study. The SBRT schedule consisted of 5 fractions of 7.25 Gy delivered every other day. Each patient was treated daily with the online adaptive plan generated according to a procedure previously published.1 Eclipse v. 16.1 was the treatment planning system used. One radiation oncologist contoured the clinical target volume (CTV) in all CBCT scans, which were acquired for all fractions and for all patients. CTV consisted of the whole prostate without seminal vesicles. Resolution of the CBCT images was 0.9 mm in left-right (L-R) and anterior-posterior (A-P) directions, and 2 mm in superior-inferior (S-I) direction..A DICOM RTSTRUCT file containing the CTV contoured in all CBCTs was exported for each patient. The delineation differences on CTVs per patient were computed using a script developed by Dupont et al.2, based on the work of Remeijer et al.3 For every anatomical direction (L-R, A- P, S-I), the script provides the mean variation (M) and the standard deviation (SD) between the five online fractions planned per patient. Delineation uncertainty is a purely systematic error ( Σ )4, and it was calculated as the SD of the means of the 20 patients. No variations of the prostate volume through the treatment course were assumed in this study. Results: The Figure shows the M values for the 20 patients. The average values of M in the superior and inferior directions were offset by more than 1 mm (1.1 and 1.4 mm, respectively). Derived systematic uncertainties ( Σ ) in prostate volume delineation were: 1.0 mm (L), 1.7 mm (R), 1.7 mm (A), 1.8 mm (P), 2.3 mm (S) and 1.7 mm (I).The largest systematic error was detected at the base of the prostate ( Σ = 2.3 mm).
the treatment plan (Plan_ref) was then reconstructed (Plan_recons) on the CBCT2 scan by placing the isocenter at the patient's position where the treatment was delivered. Original fluences and monitor units were kept. Each reconstructed plan was compared to the corresponding Plan_ref. The Table shows the evaluated metrics and the ideal dose constraints used by our group. A two-tailed t-test was performed for each metric using 0.05 as the significance level Results: 1. For all patients and fractions, no appreciable changes in lumpectomy cavity volumes were observed by the oncologist in the CBCT1 scans with respect to the pCTs.2. The Table shows the dosimetric results. Statistically significant differences were observed in the metrics analyzed for the CTV and PTV. No statistically significant differences were found for the OARs evaluated.3.The Figure shows the CTV and PTV coverage values (V95%) from all treatment fractions. Two fractions resulted in V95% < 95%, and just one reported V95% < 90%, which is the lowest dose coverage accepted in our practice.
Conclusion: According to our experience, a 6D rigid registration between the planning CT and the CBCT acquired for each treatment fraction is sufficient to ensure the CTV and OAR dosages, whenever no anatomical changes are detected in the lumpectomy area. Online adaptation does not seem to be necessary in our clinical practice. Keywords: APBI, CBCT, reconstruction Digital Poster 50 Estimation of the prostate delineation uncertainty during an online adaptive SBRT course JUAN-FRANCISCO CALVO-ORTEGA 1,2 , MARCELINO HERMIDA-LÓPEZ 3 , SANDRA MORAGUES-FEMENIA 1 , JOSÉ TORICES-CABALLERO 1 1 RADIATION ONCOLOGY, HOSPITAL QUIRÓNSALUD BARCELONA, BARCELONA, Spain. 2 RADIATION ONCOLOGY, HOSPITAL QUIRÓNSALUD MÁLAGA,
Made with FlippingBook - Share PDF online