ESTRO 2025 - Abstract Book

S2755

Physics - Dose prediction, optimisation and applications of photon and electron planning

ESTRO 2025

The calculated data are presented in Table 1.

Dose, max DHDR%/Dvmat%

Dose, min DHDR%/Dvmat%

Homoge neity D5%/D95%

Conformity V95%/Vptv

Planning time

Treatment delivery time

Number of fractions 0.4; 0.4; 0.4

Dose to Bone

1.33; 1.20; 1.25

0.95; 0.93; 0.90

0.6; 0.7; 0.5

2.4; 3.1; 2.7

15; 10; 8

4.5; 4.2; 4

3; 3.2; 2.5

Conclusion: High treatment response rates were obtained in both radiotherapy methods. Brachytherapy is more time consuming process.

VMAT demonstrated dosimetric advantages over brachytherapy technique for treatment large volumes with significant curvature. 3D printed custom bolus is a good solution to achieve desired dose distribution in complicated regions and shows good results in terms of treatment response.

Keywords: Radiotherapy, extremities, skin cancer

1257

Digital Poster Stochastic Robust Optimisation in Hyperthermia Accounting for Thermal and Dielectrical Tissue-Property Uncertainties using Polynomial Chaos Expansion Timoteo D Herrera 1,2 , Jort Groen 1,2 , Johannes Crezee 1,2 , Petra Kok 1,2 1 Radiation Oncology, Amsterdam UMC, Amsterdam, Netherlands. 2 Cancer biology and immunology, Treatment and quality of life, Cancer Center Amsterdam, Amsterdam, Netherlands Purpose/Objective: Hyperthermia, heating tumours to 40-43°C for one hour, is a powerful radiosensitizer, showing a strong dose-effect relationship that increases with achieved tumour temperature 1 . Temperature optimisation in hyperthermia treatment planning (HTP) aims at maximising tumour temperature (typically T90; the temperature reached in at least 90% of the tumour), with hard normal tissue temperature constraints. However, uncertainties in tissue/perfusion properties pose a challenge to precise temperature control, often requiring device adjustments during treatment and potentially compromising the combined effectiveness of hyperthermia and radiotherapy (thermoradiotherapy). To address this, we developed robust optimization strategies to enhance the reliability of temperature optimization by minimizing the impact of tissue/perfusion uncertainties. Material/Methods: Plan2Heat uses a temperature matrix (T-matrix) framework 2 to compute temperatures based on average values for perfusion, thermal and dielectric properties. We extended this framework by incorporating stochastic Polynomial Chaos Expansion (PCE) models to address variability 3 . We generated an average T-matrix (T avg ), from 10.000 PCE samples with different tissue properties, and a covariance matrix C, allowing to efficiently calculate, during optimisation, the average temperature and standard deviation by super-positioning. Our research implemented three new optimisation strategies:

(1) T avg 90 maximisation, with hard constraint on normal tissue temperature (max(T tissue )), (2) T avg 90 maximisation, with hard constraint on normal tissue temperature variation, and (3) combined T avg 90 maximisation and variation minimisation, with a hard constraint on max(T tissue ).