S2206
Physics - Intra-fraction motion management and real-time adaptive radiotherapy
ESTRO 2026
simulated X-rays of the same 4DCT phase [2]. The registration accuracy was quantified by calculating the absolute distance between the original GTV's centre of mass (clinically-approved structure) and the corresponding centre of mass after applying the registration’s transformation. Results: The Correlation Object's volume exhibited a non-linear increase with tolerance threshold (Figure 1). Figures 2a and 2b show the maximum possible tolerance threshold across all patients for inspiration and expiration phases. The median (IQR) absolute distances at 0mm tolerance were 2.1 (1.2-4.1) mm and 1.7 (0.9-3.8) mm for inspiration and expiration phases. Increasing the tolerance threshold to 5mm (Correlation Object volume on average 65±18% of ipsilateral lung volume) resulted in decreased median
(IQR) absolute distances of 0.3 (0.2-0.3) mm (inspiration, Figure 2c) and 0.2 (0.1-0.3) mm (expiration, Figure 2d).
Conclusion: Adding bronchial structures yielded improved accuracy compared to GTV-only registration when registering to the same 4DCT phase. The results show the importance of selecting optimal tolerance thresholds. The translation of these findings to an applied clinical setting with different anatomy and consequences on the margin concept needs further
investigation. References:
[1] Gevaert et al. (2023) PO-1883Markerless SBRT for lung treatment using ExacTrac Dynamic: a feasibility study.https://doi.org/10.1016/S0167-8140(23)66798- 2[2] Jin et al. (2008) Use of the BrainLAB ExacTrac X- Ray 6D System in Image-Guided Radiotherapy.https://doi.org/10.1016/j.meddos.2008.0 2.005 Keywords: Correlation Objects, X-ray Tumour Localization
Made with FlippingBook - Share PDF online