S2007
Physics - Dose prediction/calculation, optimisation and applications for photon and electron planning
ESTRO 2026
identify targets (PTV and CTV/ITV) and quantify the extent of PTV coverage compromise by generating expanded hypothetical PTVs (hPTVs). We present a fast and reliable tool to efficiently analyze large datasets, enabling easy assessment of the dosimetric consequences of target reduction and addressing the potential impact on microscopic extension. Material/Methods: A DICOM-based workflow was developed in Python3 comprising three sequential modules: (i) “Analysis”: identifies target volumes using structure names and dose coverage, calculates CTV/ITV-to-PTV spatial expansion, and detects cropped targets by accessing CTV/ITV-PTV inter-surface distances; (ii) “PTV reconstruction”: hPTVs of cropped targets are generated by expanding the CTV/ITV according to the previously derived expansion map; (iii) “Coverage evaluation”: computes V95, V90, D95, D90, D99, and the Coverage Conformity Index (CCI = D99/prescription) for the structures of interest. An exploratory correlation analysis was performed between dosimetric indices obtained on a testing cohort and clinical outcomes using Kaplan-Meier curves and univariate Cox regressions. Results: The workflow was validated on 95 SBRT/SABR plans for oligometastatic patients, randomly selected from the internal database, covering multiple anatomical sites and prescriptions. The CTV/ITV-to-PTV expansion was ≤ 5 mm in 79% of cases. PTV cropping occurred in 25% of cases, with a crop distance of 1-5 mm in 29% of cases, 6-10 mm in 38%, and >10 mm in 33%. Under- coverage (CCI < 0.9) was observed in 22% of treatments. Disease progression was observed in 33% of patients. Table 1 reports the statistics of the dosimetric evaluations, calculated CCIs and clinical follow-up data. Figure 1 shows the results of Progression Free Survival (PFS) stratified by both PTV cropping (yes/no) and CCI values. In both cases, the calculated Kaplan-Meier curves showed no statistical differences with p-values of 0.47 and 0.16, respectively. Univariate Cox regression analysis between PFS and the dosimetric variables evaluated in this study also revealed no significant correlations (p- values > 0.2).
Conclusion: An automated DICOM-based workflow for quantifying PTV underdosage in SBRT/SABR was successfully developed and tested. The pipeline accurately identifies targets and reconstructs hPTVs, enabling efficient large-scale analyses with minimal user intervention. Preliminary attempts to correlate dosimetric indicators with PFS did not yield significant associations, likely due to the limited sample size. Future work will focus on multicentric validation with larger datasets. References: [1] Olson, R., Mathews, L., Liu, M. et al. Stereotactic ablative radiotherapy for the comprehensive treatment of 1-3 Oligometastatic tumors (SABR- COMET-3): study protocol for a randomized phase III
Made with FlippingBook - Share PDF online