Simulation and experimental verification of ambient neutron doses in a pencil beam scanning proton therapy room as a function of treatment plan parameters

Out-of-field patient doses in proton therapy are dominated by neutrons. Currently, they are not taken into account by treatment planning systems. There is an increasing need to include out-of-field doses in the dose calculation, especially when treating children, pregnant patients, and patients with implants. In response to this demand, this work presents the first steps towards a tool for the prediction of out-of-field neutron doses in pencil beam scanning proton therapy facilities. As a first step, a general Monte Carlo radiation transport model for simulation of out-of-field neutron doses was set up and successfully verified by comparison of simulated and measured ambient neutron dose equivalent and neutron fluence energy spectra around a solid water phantom irradiated with a variation of different treatment plan parameters. Simulations with the verified model enabled a detailed study of the variation of the neutron ambient dose equivalent with field size, range, modulation width, use of a range shifter, and position inside the treatment room. For future work, it is planned to use this verified model to simulate out-of-field neutron doses inside the phantom and to verify the simulation results by comparison with previous in-phantom measurement campaigns. Eventually, these verified simulations will be used to build a library and a corresponding tool to allow assessment of out-of-field neutron doses at pencil beam scanning proton therapy facilities.

Range-shifter effects on the stray field in proton therapy measured with the variance–covariance method

Measurements in the stray radiation field from a proton therapy pencil beam at energies 70 and 146 MeV were performed using microdosimetric tissue-equivalent proportional counters (TEPCs). The detector volumes were filled with a propane-based tissue-equivalent gas at low pressure simulating a mean chord length of 2 μm in tissue. Investigations were performed with and without a beam range shifter, and with different air gaps between the range shifter and a solid water phantom. The absorbed dose, the dose-mean lineal energy, and the dose equivalent were determined for different detector positions using the variance–covariance method. The influence from beam energy, detector- and range-shifter positions on absorbed dose, LET, and dose equivalent were investigated. Monte Carlo simulations of the fluence, detector response, and absorbed dose contribution from different particles were performed with MCNP 6.2. The simulated dose response for protons, neutrons, and photons were compared with, and showed good agreement with, previously published experimental data. The simulations also showed that the TEPC absorbed dose agrees well with the ambient absorbed dose for neutron energies above 20 MeV. The results illustrate that changes in both dose and LET variations in the stray radiation field can be identified from TEPC measurements using the variance–covariance method. The results are in line with the changes seen in the simulated relative dose contributions from different particles associated with different proton energies and range-shifter settings. It is shown that the proton contribution scattered directly from the range shifter dominates in some situations, and although the LET of the radiation is decreased, the ambient dose equivalent is increased up to a factor of 3.

Dose-mean lineal energy values for electrons by different Monte Carlo codes: Consequences for estimates of radiation quality in photon beams

BACKGROUND

The microdosimetric quantity lineal energy and its mean values have proven useful for quantifying radiation quality in many situations. The ratio of dose-mean lineal energies is perhaps the simplest quantity for quantifying differences between two radiation qualities. However, published dose-mean lineal energy values from different codes may differ significantly with potential influence on radiation quality estimates.

PURPOSE

The purpose was to compare dose-mean lineal energy values from different track-structure data sets for condensed water vapor and liquid water, and to evaluate the influence on radiation quality estimations for some photon sources.

METHODS

Published dose-mean lineal energy values for 0.1 keV to 1 MeV electrons in spheres with diameters 2 nm to 1 μm, calculated with water vapor and liquid water track structure codes and proximity functions, were collected, analyzed, and compared. Data for cylinders were converted to spheres using a theoretical transformation published by Kellerer. A new set of dose-mean lineal energy values was calculated to cover the whole range of volumes of interest here using the GEANT4-DNA code. The influence from the differences between codes on radiation quality calculations was estimated using dose-mean lineal energy ratios for the photon sources ¹²⁵ I, ¹⁶⁹ Yb, and ¹⁹² Ir relative to ⁶⁰ Co.

RESULTS

The theoretical relation for converting the dose-mean lineal energy between different geometrical volumes, results in differences up to 10% between cylinders and spheres depending on electron energy and target size, in agreement with published simulated results. For spheres with diameter above 100 nm, dose-mean lineal energy values for condensed water vapor and liquid water are with few exceptions within ±10%. Below 100 nm, the difference increases with decreasing diameter reaching a factor of two at 2 nm. The values from water vapor codes are in general larger than from liquid water codes. If the dose-mean lineal energy ratio is based on condensed water vapor instead of liquid water, the ratio differs less than 9% for the nuclides ¹²⁵ I, ¹⁶⁹ Yb, and ¹⁹² Ir relative to ⁶⁰ Co independent of the volume simulated. However, a specific value of the dose-mean lineal energy ratio, is found at a larger target diameter in liquid water than in condensed water vapor.

CONCLUSIONS

When ratios of the dose-mean lineal energy are used as a measure of the radiation quality it is important to compare values for geometrically equal target shapes. A practical method of converting values for cylinders of equal diameter and height to spheres was demonstrated. Although dose-mean lineal energy values calculated with water vapor and liquid water codes may differ significantly, the radiation quality, in terms of ratios of dose-mean lineal energy, for the three photon sources ¹⁹² Ir, ¹⁶⁹ Yb, and ¹²⁵ I relative to ⁶⁰ Co, agree within 9%. The same ratio appears at a larger diameter when a liquid water code is used. It is therefore important to use the same code in radiation quality investigations. The present findings may be of special interest in studies related to the relative biological effectiveness (RBE).

RADIATION PROTECTION MEASUREMENTS WITH THE VARIANCE-COVARIANCE METHOD IN THE STRAY RADIATION FIELDS FROM PHOTON AND PROTON THERAPY FACILITIES

The microdosimetric variance-covariance method was used to study the stray radiation fields from the photon therapy facility at the Technical University of Denmark and the scanned proton therapy beam at the Skandion Clinic in Uppsala, Sweden. Two TEPCs were used to determine the absorbed dose, the dose-average lineal energy, the dose-average quality factor and the dose equivalent. The neutron component measured by the detectors at the proton beam was studied through Monte Carlo simulations using the code MCNP6. In the photon beam the stray absorbed dose ranged between 0.3 and 2.4 μGy per monitor unit, and the dose equivalent between 0.4 and 9 μSv per monitor unit, depending on beam energy and measurement position. In the proton beam the stray absorbed dose ranged between 3 and 135 μGy per prescribed Gy, depending on detector position and primary proton energy.

Comment on 'Monte Carlo calculated microdosimetric spread for cell nucleus-sized targets exposed to brachytherapy (125)I and (192)Ir sources and (60)Co cell irradiation'

The relative standard deviation, σr,D, of calculated multi-event distributions of specific energy for (60)Co ϒ rays was reported by the authors F Villegas, N Tilly and A Ahnesjö (Phys. Med. Biol. 58 6149-62). The calculations were made with an upgraded version of the Monte Carlo code PENELOPE. When the results were compared to results derived from experiments with the variance method and simulated tissue equivalent volumes in the micrometre range a difference of about 50% was found. Villegas et al suggest wall-effects as the likely explanation for the difference. In this comment we review some publications on wall-effects and conclude that wall-effects are not a likely explanation.

Determinations of H(10) and its dose components onboard aircraft

Aircrew is in general receiving a higher average annual dose than other occupationally exposed personnel, and about half of the effective dose is deposited by high-LET neutron secondaries. A recent investigation of the cancer incidence following the atomic bombs at Hiroshima and Nagasaki has put forward the possibility that the relative biological efficiency for neutrons could be underestimated. If so, the effective dose to aircrew from this component would increase and the estimation of this component will become even more important. Different ambient dose equivalent measurement techniques and calculation methods have recently been compared on a dedicated flight. The experimental results are compared with calculations made with the codes EPCARD 3.2 and an updated version of FLUKA and different galactic proton spectra. The aircraft circulated within the target areas at two constant altitudes with a flight route variation of only about 1 degrees in longitude and latitude to reduce the influence from variations in atmospheric and geomagnetic shielding. The instrumentation consisted of tissue-equivalent proportional counters (TEPC) and a silicon diode spectrometer. Measurements were performed for 2 h to reduce the statistical uncertainties in the results. The TEPCs were evaluated either according to single-event analysis techniques or the variance-covariance method. Besides the total ambient dose equivalent, the instruments can be evaluated to reveal the low- and high-LET components. The EPCARD and FLUKA simulations can determine the contribution from each type of particle directly. The ratio between the calculated and the measured average value of the ambient dose equivalent rate was 1.00 +/- 0.08 with all instruments included for EPCARD and 0.97 +/- 0.07 when FLUKA was used. The measured high-LET component and the calculated neutron component are not quite identical, but should be similar. The agreement was always within 20%. The high-LET component contributed with about 57% at N57 E8 and 48% at N42 E12.