Abstract
Introduction
Monte Carlo simulations are the gold standard for radiation dosimetry, but developing accurate models for kilovoltage cone-beam computed tomography (kV-CBCT) systems can be challenging due to non-availability of vendor-supplied geometry information. This study presents the results of Monte Carlo modeling of the kV-CBCT beamline of the Halcyon 2.0 linear accelerator using semi-empirical methods to derive equivalent source and filtration models.
Method
Equivalent energy spectra for 100, 125, and 140 kV beams were determined by matching measured beam characteristics (half-value layers and air kerma outputs) using SPEKTR 3.0 software. The bowtie filter profile was reconstructed from transmission measurements, enabling generation of a 3D filter model. These models were incorporated into a Geant4/GATE Monte Carlo simulation. Computed percentage depth doses (PDDs) and off-axis profiles were benchmarked against water phantom measurements.
Results
Maximum differences between measured and computed PDDs were 3.47% (100 kV), 3.65% (125 kV), and 3.27% (140 kV). For off-axis profiles, maximum differences were 3.73% (100 kV), 6.84% (125 kV), and 4.44% (140 kV) within the central beam. Larger discrepancies up to 22% occurred in high-gradient penumbral regions due to mismatches in spatial resolution between detectors and Monte Carlo scoring geometry.
Conclusion
This study presents the first comprehensive Monte Carlo model of the Halcyon 2.0 kV-CBCT system using measurement-derived equivalent models. The good agreement with experimental data validates the accuracy of the models, which can enable imaging dose calculations and other applications for this system.