Monte Carlo calculated ionization chamber correction factors in clinical proton beams - deriving uncertainties from...

For the update of the IAEA TRS-398 Code of Practice (CoP), global ionization chamber factors (f_Q) and beam quality correction factors (k_Q) for air-filled ionization chambers in clinical proton beams have been calculated with different Monte Carlo codes. In this study, average Monte Carlo calculated f_Q and k_Q factors are provided and the uncertainty of these factors is estimated. Average f_Q factors in monoenergetic proton beams with energies between 60 MeV and 250 MeV were derived from Monte Carlo calculated f_Q factors published in the literature. Altogether, 195 f_Q factors for six plane-parallel and three cylindrical ionization chambers calculated with penh, fluka and geant4 were incorporated. Additionally, a weighted standard deviation of f_Q factors was calculated, where the same weight was assigned to each Monte Carlo code. From average f_Q factors, k_Q factors were derived and compared to the values from the IAEA TRS-398 CoP published in 2000 as well as to the values of the upcoming version. Average Monte Carlo calculated f_Q factors are constant within 0.6% over the energy range investigated. In general, the different Monte Carlo codes agree within 1% for low energies and show larger differences up to 2% for high energies. As a result, the standard deviation of f_Q factors increases with energy and is ∼0.3% for low energies and ∼0.8% for high energies. k_Q factors derived from average Monte Carlo calculated f_Q factors differ from the values presented in the IAEA TRS-398 CoP by up to 2.4%. The overall estimated uncertainty of Monte Carlo calculated k_Q factors is ∼0.5%-1% smaller than the uncertainties estimated in IAEA TRS-398 CoP since the individual ionization chamber characteristics (e.g. fluence perturbations) are considered in detail in Monte Carlo calculations. The agreement between Monte Carlo calculated k_Q factors and the values of the upcoming version of IAEA TRS-398 CoP is better with deviations smaller than 1%.