Validation of Commercial Monte Carlo Secondary Dose Calculation Using External Audit End-to-End Phantoms
Abstract
Purpose
Independent verification of IROC phantom credentialing results is valuable for institutional quality assurance and confidence in treatment planning system performance. This work aimed to verify the accuracy of a commercially available Monte Carlo (MC) dose algorithm using end-to-end IROC measurement data across multiple linear accelerators and phantom types.
Methods
Three brain, three head and neck (HN), and six lung IROC phantoms were irradiated. These end-to-end tests were evaluated using IROC reports generated by the Radiation Quality Assurance Lab with 7%/5mm gamma criteria. Dose profile data along the superior–inferior, right–left, and anterior–posterior axes were extracted from IROC report plots using an in-house software. Corresponding dose profiles were recalculated using a commercially available secondary dose calculation engine (RadMonteCarlo, Radformation, NY) and sampled through phantom centers using matched orientations. Profiles were centered and aligned to each film’s coordinate system before comparison. Gamma pass rate agreement between IROC-extracted and MC-extracted profiles was evaluated using IROC criteria, and more clinically relevant criteria: 5%/3mm for all sites, 7%/4mm for HN, and 5%/1mm for radiosurgery brain cases. Thermoluminescent dosimeter (TLD) measurements were also analyzed, with percent dose differences evaluated relative to prescription using a 7% acceptance criterion.
Results
All phantom cases that passed IROC-reported criteria also passed MC-based gamma analysis, with pass rates ranging from 97.6-100%. For 5%/3mm criteria, pass rates ranged from 89.5-99% for brain and HN cases, and 90.3-98.3% for lung cases. TLD measurements demonstrated good agreement for both target and OAR, with target differences within 4% and OARs within 3%. No TLD measurement exceeded the IROC acceptance criterion.
Conclusion
IROC-reported gamma analysis results were reproducible and consistent with independently extracted MC dose profiles across multiple phantom types and anatomical sites. These findings support the robustness of current IROC gamma criteria and demonstrate the accuracy of independent MC-based calculation across multiple sites and machines.