gms | German Medical Science

Introduction: a new proton gantry dedicated to advance scanning has been build at Paul Scherrer Institute (PSI) and is under commissioning. Patient positioning is foreseen to occur directly in the treatment room with an in-room CT scanner. In addition, the gantry will be equipped with an X-ray system that will allow acquiring images parallel to the proton beam and simultaneously with proton delivery. Secondary neutron dose produced during beam delivery may damage the hardware of our imaging equipments. Of particular concern is the neutron dose to the flat panel distal to the patient. The purpose of this study is to estimate the secondary neutron dose at different positions in the treatment room, which are critical for the flat panel and the CT scanner.

Methods and materials: Measurements of the secondary neutron dose are performed during irradiation of a water phantom with a 200 MeV proton pencil beam using two commercial available neutron counters, i.e. the LB6411 from Berthold and the WENDI-2 from Thermo, at three main locations in the treatment room, that is, just distal to the water phantom (flat panel position), at about 5 m from the beam axis (CT position) and at about 3 m distal to the phantom. The response of the WENDI-2 detector compared to the LB6411 is superior in the region of high energy neutrons (above 20 MeV). For comparison Monte Carlo simulations (MCNPX) are performed to calculate the energy spectrum of the secondary neutrons and the neutron dose.

Results: The simulations show that the secondary neutron spectrum distal to the patient and along the beam axis (0°) is significantly different than the one lateral (at 90°). Distal to the phantom the spectrum is dominated by high energy neutrons (above 20 MeV) whereas at 90° the spectrum is dominated by low energy neutrons. Distally, the LB6411 measurements have been systematically lower than those obtained with the WENDI-2 (in average 50% lower). The WENDI-2 delivers more realistic results for dose measurements along the beam axis since it is more sensitive to high energy neutrons. The dose calculated by the simulation deviates within 20% of the measurements obtained with the WENDI-2. The measured dose just distal to the phantom is 8.2x10^-15 Sv/proton (i.e., 0.0008 Sv per treatment Gy), at 5 m laterally is reduced by a factor of ~300 and at 3 m distally by a factor of ~40.

Conclusions: Distal to the phantom, the expected main contribution of secondary neutrons has energies greater than 20 MeV. Because of the low response of the LB6411, the measured dose may be underestimated by up to a factor of two. More realistic doses can be obtained with the WENDI-2. The good agreement with the simulations, in which we intentionally neglect the interaction of the protons with the components of the nozzle, confirms that, during active scanning, secondary neutron dose is only produced by the interaction of the primary proton beam with the patient.