S1951
Physics - Dose prediction/calculation, optimisation and applications for photon and electron planning
ESTRO 2026
mean deviation of D(Vcrit) due to uncertainties. Radio- sensitivity variations can be added. Key benefit of MD,OAR-values is their direct relation to clinical outcomes through the NTCP. Here, we investigate whether the approach can be applied for CTVs, and how PTV margins can be included. Material/Methods: In Fig.1, MD,OAR is visualized, and a similar approach indicated for a CTV. Treating the CTV with D50 would cure the average patient with the average treatment, but MD,CTV is required to account for interpatient sensitivity and dosimetric variations and obtain a high TCP. Since PTV-margins MPTV account more effectively for geometric variations, we actually treat our patients with two kinds of margin. To study their interplay, we simulated geometric and dosimetric errors ( Σ and σ ) in the TPS for CTVs with different MPTV, obtained SD(Dcrit) and MD(Dcrit), and determined the relation between MPTV and MD,CTV. For practical reasons, we simulated motions of the actual CTV, treated with clinical MPTV, and the PTV itself, representing the case with margin zero. This allows us to calculate the effect of MPTV on MD,CTV.
level varied between 61% and 87% (median = 78%).The Gradient Index ranged from 2.25 to 8.05 (median 5.28), while the Paddick Conformity Index ranged from 0.61 to 1.20 (median 0.76). The two-parameter model achieved A = 4.85 ± 0.09, B = 0.26 ± 0.02, with MSE = 0.80, MAE = 0.67, and R ² = 0.63. Residual analysis showed a Gaussian distribution centered at zero, confirming the absence of systematic bias and supporting the model’s goodness of fit (Figure 1). The final empirical relationship GI = 4.85/PTV0.26 is consistent with previously reported trends [Ballangrud
et al., 2018]. Conclusion:
A simple empirical model accurately predicts the Gradient Index from the Planning Target Volume, providing a fast and interpretable tool for plan quality evaluation in stereotactic radiotherapy. The model can support automated script-based plan verification and contribute to standardization of dosimetric criteria in clinical practice. Future work will extend validation to multi-institutional datasets and integrate additional dosimetric parameters. References: [1] Ballangrud Å, Kuo LC, Happersett L, Lim SB, Beal K, Yamada Y, Hunt M, Mechalakos J. Institutional experience with SRS VMAT planning for multiple cranial metastases. J Appl Clin Med Phys. 2018 Mar;19(2):176–183. doi: 10.1002/acm2.12284. Epub 2018 Feb 23. PMID: 29476588; PMCID: PMC5849827. Keywords: Stereotactic Radiotherapy, Gradient Index, VMAT Digital Poster 3790 Combining dosimetric and geometric safety margins for CTVs Joep Stroom 1 , Sandra Vieira 2 , Carlo Greco 2 1 Radiation Oncology, Champalimaud Foundation, Lisboa, Portugal. 2 Radiation Oncology, Champalimaud Foundation, Lisbon, Portugal Purpose/Objective: Previously1, the relation between dose constraints, safety margins, and treatment uncertainties for OARs was established. Delivered doses are spread around the planned dose due to geometric and dosimetric uncertainties. Assuming that the highest doses are responsible for the toxicities related to the dose constraints, a dosimetric margin (MD,OAR) recipe was derived: MD,OAR = Dtrue – Dcrit = Φ -1(1- NTCP) * SD(Dcrit) + MD(Dcrit), Eq.1with Dtrue the real, fixed dose level above which toxicities occur that is equal to D50, (Dcrit , Vcrit) the dose constraint, Φ -1 the inverse cumulative normal distribution, NTCP the toxicity probability associated with the constraint, and SD and MD the standard- and
Results: We used 192 CTVs planned with D99%>95%, 80 with hypofractionation, and 112 with regular RT. The difference between MD,CTV calculated with Eq.1 between MPTV=0 and the clinical MPTV are shown in Fig.2 for three realistic MPTV recipes. Without MPTV, more dose is required to maintain TCP. The larger the
Made with FlippingBook - Share PDF online