ABSTRACT
Inside an IBA proton therapy centre, secondary neutrons are produced due to nuclear interactions of the proton beam with matter mainly inside the cyclotron, the beam line, the treatment nozzle and the patient. Accurate measurements of the neutron ambient dose equivalent H*(10) in such a facility require the use of a detector that has a good sensitivity for neutrons ranging from thermal energies up to 230 MeV, such as for instance the WENDI-2 detector. WENDI-2 measurements have been performed at the Westdeutsches Protonentherapiezentrum Essen, at several positions around the cyclotron room and around a gantry treatment room operated in two different beam delivery modes: Pencil Beam Scanning and Double Scattering. These measurements are compared with Monte Carlo simulation results for the neutron H*(10) obtained with MCNPX 2.5.0 and GEANT4 9.6.
Subject(s)
Neutrons , Proton Therapy/methods , Radiometry/instrumentation , Radiometry/methods , Algorithms , Calibration , Computer Simulation , Cyclotrons , Facility Design and Construction , Germany , Humans , Monte Carlo Method , Protons , Radiation Dosage , Radiotherapy Dosage , Reproducibility of Results , Scattering, RadiationABSTRACT
The WENDI-II rem meter is one of the most popular neutron dosemeters used to assess a useful quantity of radiation protection, namely the ambient dose equivalent. This is due to its high sensitivity and its energy response that approximately follows the conversion function between neutron fluence and ambient dose equivalent in the range of thermal to 5 GeV. The simulation of the WENDI-II response function with the Geant4 toolkit is then perfectly suited to compare low- and high-energy hadronic models provided by this Monte Carlo code. The results showed that the thermal treatment of hydrogen in polyethylene for neutron <4 eV has a great influence over the whole detector range. Above 19 MeV, both Bertini Cascade and Binary Cascade models show a good correlation with the results found in the literature, while low-energy parameterised models are not suitable for this application.