Out-of-field dose assessment in electron beam radiotherapy: experimental measurements, GEANT4/GATE Monte Carlo simulations, and evaluation of the Eclipse GGPB algorithm

Abstract The aim of this study was to quantitatively evaluate out-of-field (peripheral) doses in electron beam radiotherapy, to compare experimental measurements with treatment planning system (TPS) calculations and Monte Carlo (MC) simulations, and to discuss their clinical relevance. Peripheral dose measurements were performed in a water phantom using a Roos parallel-plate ionization chamber for 6, 9, 12, and 15 MeV electron beams delivered by a Varian linear accelerator with various applicator sizes. Under identical conditions, dose distributions were recalculated with the Eclipse TPS using the General Gaussian Pencil Beam (GGPB) algorithm and with Geant4/GATE MC simulations. The dependence of peripheral dose on energy, applicator size, and distance from the field edge was assessed. Experimental data showed that peripheral doses increased with both beam energy and applicator size, while decreasing with greater depth and distance from the field border. TPS calculations consistently underestimated peripheral doses, with discrepancies up to 30% at higher energies and near-field regions. By contrast, MC simulations demonstrated excellent agreement with measurements across all energies and field sizes, with differences typically within 1%–2%. The results demonstrate that the GGPB algorithm tested in this study substantially underestimates out-of-field doses. Experimental dosimetry and validated MC simulations are essential for comprehensive dose evaluation, particularly for estimating secondary cancer risks, protecting adjacent critical organs, and ensuring the safety of implantable devices. The integration of GPU-accelerated Monte Carlo algorithms and AI-based prediction models into clinical workflows may further improve the accuracy and efficiency of peripheral dose calculations, ultimately enhancing patient safety.