Monte Carlo Simulations of Neutron Ambient Dose Equivalent in a Novel Proton Therapy Facility Design

Abstract

Purpose: The neutron shielding properties of the concrete structures of a proposed proton therapy facility were evaluated with help of the Monte Carlo technique. The planned facility's design omits the typical maze-structured entrances to the treatment rooms to facilitate more efficient access and, instead, proposes the use of massive concrete/steel doors. Furthermore, straight conduits in the treatment room walls were used in the design of the facility, necessitating a detailed investigation of the neutron radiation outside the rooms to determine if the design can be applied without violating existing radiation protection regulations. This study was performed to investigate whether the operation of a proton therapy unit using such a facility design will be in compliance with radiation protection requirements.

Methods: A detailed model of the planned proton therapy expansion project of the University of Texas, M. D. Anderson Cancer Center in Houston, Texas, was produced to simulate secondary neutron production from clinical proton beams using the MCNPX Monte Carlo radiation transport code. Neutron spectral fluences were collected at locations of interest and converted to ambient dose equivalents using an in-house code based on fluence to dose-conversion factors provided by the International Commission on Radiological Protection.

Results and conclusions: At all investigated locations of interest, the ambient dose equivalent values were below the occupational dose limits and the dose limits for individual members of the public. The impact of straight conduits was negligible because their location and orientation were such that no line of sight to the neutron sources (ie, the isocenter locations) was established. Finally, the treatment room doors were specially designed to provide spatial efficiency and, compared with traditional maze designs, showed that while it would be possible to achieve a lower neutron ambient dose equivalent with a maze, the increased spatial (and financial) requirements may offset this advantage.

Keywords: Monte Carlo study; neutron shielding; proton therapy.

Figure 1.

The facility layout of the Monte Carlo model, including some relevant neutron spectral fluence receptor locations. Receptors 17, 20, 23, and 26 are located in the treatment control rooms, and receptors 28, 29, and 30 are located in the main control room. Additional local shielding was modeled at location A. The lower image depicts a schematic top-down view of the treatment room door designs showing the sandwich design and the 9.4-cm air gap between the door and the walls of the sliding door (a) and the more traditional maze designs (b and c). The filled circles are the isocenter location in the rooms, and the arrows indicate 2 out of 4 simulated proton beam directions. The remaining 2 are normal and antinormal to the image plane.

Figure 2.

The fluence to dose conversion factor, h, in units of pSv cm^–2, taken from ICRP 116 [19] and from the in-house conversion code (F2D). The solid line shows the tabulated data from ICRP 116, and the scatterplot (circles) depicts values calculated with F2D. Note that thermal neutrons convert to rather small doses, and only neutrons with kinetic energies larger than about 100 keV contribute significantly to the ambient dose equivalent.

Figure 3.

Panel (a) shows the 2-dimensional logarithmic plot of total neutron fluence in arbitrary units, weighted with the annual beam-loss estimates, in the isocenter plane. The lower plots show neutron spectral fluence Φ per cm² and source proton, p, from receptors located at the center of the accelerator, at the entrance and exit of the maze to the accelerator room in semi-log (b) and in log-log (c) plots.

Figure 4.

Ambient dose equivalent, H*(10) in units of mSv per year, based on summation over the ambient dose equivalent spectra weighted with annual (a) and hourly (b) proton loss estimates. All receptors, except the ones inside the accelerator room and inside the treatment room, are shown here. The hourly H*(10) data point for TXroom describes the value in the treatment room next to the accelerator when the beam was delivered to the treatment room next to it. Panel (c) shows the frequency f of the differences in annual ambient neutron dose equivalent between simulations of the complete facility with and without all the conduits in place. Panel (d) shows relative neutron ambient dose equivalent values simulated at the exit of a treatment room featuring different door designs.