See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/277958335

PHITS simulations of absorbed dose out-of-field and neutron energy spectra

for ELEKTA SL25 medical linear accelerator

Article in Physics in Medicine & Biology · June 2015

DOI: 10.1088/0031-9155/60/12/N261 · Source: PubMed

CITATIONS READS

14 2,107

2 authors, including:

Lembit Sihver
Chalmers University of Technology
330 PUBLICATIONS 5,792 CITATIONS

SEE PROFILE

All content following this page was uploaded by Monika Puchalska on 08 April 2016.

The user has requested enhancement of the downloaded file.

Home Search Collections Journals About Contact us My IOPscience

PHITS simulations of absorbed dose out-of-field and neutron energy spectra for ELEKTA

SL25 medical linear accelerator

This content has been downloaded from IOPscience. Please scroll down to see the full text.

2015 Phys. Med. Biol. 60 N261

(http://iopscience.iop.org/0031-9155/60/12/N261)

View the table of contents for this issue, or go to the journal homepage for more

Download details:

IP Address: 128.131.238.233
This content was downloaded on 12/06/2015 at 13:45

Please note that terms and conditions apply.

 Institute of Physics and Engineering in Medicine Physics in Medicine & Biology

Phys. Med. Biol. 60 (2015) N261–270 doi:10.1088/0031-9155/60/12/N261

Note

PHITS simulations of absorbed dose

out-of-field and neutron energy spectra for
ELEKTA SL25 medical linear accelerator
Monika Puchalska and Lembit Sihver1
Applied Physics, Chalmers University of Technology, 41296 Gothenburg, Sweden

E-mail: monika.puchalska@tuwien.ac.at

Received 24 March 2015, revised 22 April 2015

Accepted for publication 23 April 2015
Published 9 June 2015

Abstract
Monte Carlo (MC) based calculation methods for modeling photon and
particle transport, have several potential applications in radiotherapy. An
essential requirement for successful radiation therapy is that the discrepancies
between dose distributions calculated at the treatment planning stage and those
delivered to the patient are minimized. It is also essential to minimize the dose
to radiosensitive and critical organs. With MC technique, the dose distributions
from both the primary and scattered photons can be calculated. The out-of-field
radiation doses are of particular concern when high energy photons are used, since
then neutrons are produced both in the accelerator head and inside the patients.
Using MC technique, the created photons and particles can be followed and the
transport and energy deposition in all the tissues of the patient can be estimated.
This is of great importance during pediatric treatments when minimizing the
risk for normal healthy tissue, e.g. secondary cancer. The purpose of this work
was to evaluate 3D general purpose PHITS MC code efficiency as an alternative
approach for photon beam specification. In this study, we developed a model of
an ELEKTA SL25 accelerator and used the transport code PHITS for calculating
the total absorbed dose and the neutron energy spectra infield and outside the
treatment field. This model was validated against measurements performed with
bubble detector spectrometers and Boner sphere for 18 MV linacs, including both
photons and neutrons. The average absolute difference between the calculated
and measured absorbed dose for the out-of-field region was around 11%. Taking
into account a simplification for simulated geometry, which does not include any
potential scattering materials around, the obtained result is very satisfactorily.
A good agreement between the simulated and measured neutron energy spectra
was observed while comparing to data found in the literature.
1
Author towhom any correspondence should be addressed. Current address: Institute of Atomic and Subatomic
Physics, Vienna University of Technology, 1020 Vienna, Austria.

0031-9155/15/12N261+10$33.00 © 2015 Institute of Physics and Engineering in Medicine Printed in the UK N261
Phys. Med. Biol. 60 (2015) N261 Note

Keywords: Monte Carlo, ELEKTA linac, neutron, out-of-field

(Some figures may appear in colour only in the online journal)

Introduction

Monte Carlo (MC) simulations of radiation transport inside the treatment machine heads pro-
vide practical means for obtaining dose distribution, contribution to the total dose from the pri-
mary and secondary particles and energy spectra. To confirm the validity of the models included
in the MC codes, the accuracy of the simulated results must be verified to experimental data.
The PHITS code has already been benchmarked against heavy ion radiotherapy (Sato et al
2009) but not to treatment modalities involving photons where neutrons are produced by photo-
nuclear reactions, when the energy of the incident photon is higher than the threshold energy of
the (γ, n) reaction. This threshold depends on the atomic number of the target: for high atomic
numbers it is around 8 MeV while for low atomic numbers the threshold is higher. The neutron
production threshold is 16 MeV for oxygen and 18.7 MeV for carbon (Fuller 1985). Therefore
linacs with photon energies in the range of 18–25 MeV will produce undesired fast neutrons,
both in the accelerator head and directly in the bodies of the patients. However, since the peak
(γ, n) cross-sections for high Z materials are around 50 times higher than for low Z materials
(W: 400 mb; C: 8 mb), the accelerator head provides the major contribution (Zanini et al 2004).
For quantitative comparison, it can be said that the energy threshold of photoneutron produc-
tion for W and Pb are 6.74 MeV and 6.19 MeV, respectively. Whereas respective thresholds
for Cu and Fe have been reported to be 9.91 MeV and 7.65 MeV (Swanson 1978). The main
neutron producing materials, tungsten and lead, have very low absorption cross sections for the
energy range of neutron produced in linac head. Therefore, photoneutrons have a great chance
to penetrate the shielding and reach the patient and bunker walls. Even if the out-of-field dose
is several orders of magnitude below the dose to the tumor, secondary malignancies with a high
latency period can be induced especially in radiosensitive organs. While a large number of stud-
ies exist for Varian, Siemens and Philips machines, few data have been published for ELEKTA
machines. Current status of the studies is published in e.g. Takam et al (2011).
This paper is focused on PHITS simulations of 18 MV photons produced by the ELEKTA
SL25 machine at Klinikum Goethe Universität (KGU) in Frankfurt, Germany and was one of
the tasks of the European project ALLEGRO. Comparison to the measured data by Kaderka
(2011) and Kaderka et al (2012) from Helmholtzzentrum für Schwerionenforschung GmbH
(GSI) in Darmstadt, Germany and other data found in the literature was performed.

Materials and methods

PHITS code

The general purpose 3D Monte Carlo particle transport code PHITS (Sato et al 2013) can
calculate the transport of particles through any material in a wide energy range. PHITS has
proven its applicability in various fields of research e.g. space radiation dosimetry (Puchalska
et al 2012), and heavy ion radiotherapy (Sato et al 2009).
The neutron data libraries included in the PHITS ver.2.64 package, used in this study, were
compiled from JENDL-4.0 (JENDL-4.0 Updated Files). Recently, the neutron nuclear data
library based on JENDL-4.0 was revised and photo- and electro-atomic data libraries were devel-
oped based on EPDL97 (Cullen et al 1997) for photons and EEDL (Perkins et al 1991) and

N262
Phys. Med. Biol. 60 (2015) N261 Note

Figure 1. Geometry of ELEKTA SL25 linac head.

ITS3.0 (Halbleib et al 1992) for electrons. Photonuclear reactions are treated for the energy region
between 2–30 MeV in which the giant dipole resonance is the dominant reaction mechanism.
Typically, 1 billion particle histories were run for each simulation, requiring approximately
800 h of CPU time. The neutron cutoff energies (the energy below which particles are no
longer tracked) was set to 10−10 MeV. Electron and photon cutoff kinetic energies were set to
100 keV and 1 keV, respectively.

Geometry

The treatment head model includes beam line components such as the target, double flatten-
ing filter, jaws and a multi-leaf collimator (MLC). Both the jaws and the MLC were fitted to
achieve a treatment field of 5   ×   5 cm2 at the distance of 100 cm from the source. The geom-
etry was constructed according to technical information provided by Schardt and Kaderka
from GSI (personal communication 2010). Figure 1 shows the geometry of ELEKTA SL25.
Detailed information about material composition is shown in table 1.
The incident electron beam, striking a 1 mm thick tungsten alloy target attached to a 9 mm
thick copper backing plate, was simulated by a spot of 1 mm spatial spread and mean energy

N263
Phys. Med. Biol. 60 (2015) N261 Note

Table 1. Materials of the various elements of the linac head considered in this work.

ELEKTA SL25, 18 Varian Clinac, ELEKTA

MV (GSI personal 18 MV SL20i
communication; (Martínez (Ongaro et al
Martínez 2013) 2013) 2000)
Material Material Material
[%] [%] [%]
ρ[g cm−3] ρ[g cm−3] ρ[g cm−3]
Target W/Ni/Fe W W
95/3.75/1.25 100 100
18.0 19.3 19.3
Target cover Cu Cu Cu
100 100 100
8.96 8.96 8.96
Primary collimator W/Ni/Fe W W
95/3.75/1.25 100 100
18.0 19.3 19.3
Flattening filter Fe/Cr/Ni (Al)a Ta (Fe)b Fe
74/18/8 — 100
8.03 (2.73) 16.65 (7.87) 7.87
Multi-leaf collimator (MLC) W/Ni/Fe W W/Pb
95/3.75/1.25 100 —
18.0 19.3 —
Jaws W/Ni/Fe W W/Pb
95/3.75/1.25 100 —
18.0 19.3 —
a
Flattening filter made of steel with a cover of Al.
b
Flattening filter made of Ta with a cover of Fe.
Note: The percentage compositions and densities are also given.

of 18 MeV with full width half maximum (FWHM) of 2%. These values were selected on the
basis of values used previously in the literature (Mesbahi 2006).
A water phantom with dimensions of 50   ×   50   ×   50 cm3 was simulated under the treatment
head with a source–surface distance (SSD) of 100 cm. The dose was calculated for in-plane
(patient superior-inferior) and cross-plane (patient left-right) directions at the depth of maxi-
mum dose (3 cm) and at a depth of 10 cm. The neutron spectra were calculated at the surface
of the water phantom and at 10 cm depth. The results were normalized to 1 Gy in the target
or to 1 monitor unit (MU), which is defined as 1 cGy delivered in the center of the field at the
depth of maximum dose.

Results

Photons

Depth-dose profile for ELEKTA SL25 simulated by PHITS code and compared to experimen-
tal data by Kaderka (2011) is presented in figure 2. The depth-dose curves were normalized to
the integral under the distribution curve within the relevant depth range. The resulted depth-
dose curves from the Monte Carlo calculations are broken into various components. Photons
created at the target that did not interact with any part of the accelerator are referred to as
primary photons. The contribution from contaminant electrons and positrons originating from

N264
Phys. Med. Biol. 60 (2015) N261 Note

1.1
40
1.0

secondaries contr.
to total dose (%)
0.9 30

0.8 20

0.7 10
Relative dose

0.6
0
0 100 200 300 400 500
0.5
Depth in water (mm)
0.4 PHITS total
PHITS primary photons
0.3
PHITS secondaries
0.2 Kaderka (2011)

0.1

0.0

0 50 100 150 200 250 300 350 400 450 500

Depth in water (mm)

Figure 2. Comparison of Monte Carlo calculated and measured depth-dose curves

along with calculated components from primary photons and secondary electrons and
positrons originating from air and the ELEKTA treatment head. Data for a 5   ×   5 cm2
field size for infield 18 MV photons. Statistical uncertainties ranged between 0% and
approximately 5%, depending on the depth.

photons that interacted with air and the ELEKTA treatment head are referred to as secondar-
ies. Their percentage contribution is shown in the sub-plot of figure 2.
Lateral dose profiles calculated for in-plane and cross-plane directions at the depth of 3 cm
and 10 cm in water for the ELEKTA linac, together with the comparison to the experimental
data measured by Kaderka et al (2012) are presented in figure 3. All results were normalized
to the dose at the isocenter at the depths of 3 cm or 10 cm in the water phantom.

Neutrons

In figure 4, the neutron spectra are shown for an isocenter and for a location 30 cm from the iso-
center at the water surface and at the depth of 10 cm inside the water phantom for ELEKTA SL25.
Neutron spectra for the out-of-field are shown in figure 5, which compares the simulated
neutron spectra with the measured ones at the surface between 8 cm and 40 cm from the iso-
center. In table 2, the integral neutron fluence (10 keV–20 MeV) calculated in this study is
compared with the reported calculated and measured data provided by some authors in the
literature. While a large collection of neutron data is available for Varian, Siemens and Philips
linac machines (Takam et al 2011); only few studies have been conducted for ELEKTA
linac (Ongaro et al 2000, Howell et al 2009, Kaderka et al 2012). In the paper by Ongaro
et al (2000) the measurements were performed with bubble detector spectrometers (BDS) at
8 cm and 15 cm from the isocenter with 10   ×   10 cm2 field size for ELEKTA SL20i. Howell
et al (2009) used a Boner sphere to evaluate the neutron spectra at a distance of 40 cm from
the isocenter for ELEKTA Precise where jaws and MLC were closed. The most recent data
with BDS were published by Kaderka et al (2012) for ELEKTA SL25 with the filed size of
5   ×   5 cm2. Zanini et al (2004) reported data for Varian linac at 15 cm from the isocenter with

N265
Phys. Med. Biol. 60 (2015) N261 Note

cross-plane in-plane
1
3 cm depth 3 cm depth

(a)
30 (b) 30

difference (%)
15 15

difference (%)
0.1 0 0
-15 -15
-30 -30
0 50 100 150 200 0 50 100 150 200
Distance to isocenter (mm) Distance to isocenter (mm)
0.01
Relative dose

Kaderka et al. 2012

PHITS
1E-3
1
10 cm depth 10 cm depth
(c) 30
(d) 30

difference (%)
15
difference (%)

15

0.1 0 0

-15 -15

-30 -30
0 50 100 150 200 0 50 100 150 200
Distance to isocenter (mm) Distance to isocenter (mm)
0.01

Mesbahi 2006
1E-3
0 50 100 150 200 0 50 100 150 200

Distance to isocenter (mm)

Figure 3. Lateral dose profiles in the water phantom at the depth of 3 cm in cross-plane
direction (a) and in-plane direction (b); and the depth of 10 cm in cross-plane direction
(c) and in-plane direction (d), with a 5   ×   5 cm2 field of 18 MV photons calculated by
PHITS and measured by Kaderka et al (2012) and Mesbahi (2006) for ELEKTA SL25.

the jaws of 40   ×   40 cm2 and MLC of 10   ×   10 cm2. Kry et al (2007) presented data for Varian
machine at 40 cm from the isocenter for the filed size of 3.6   ×   4 cm2 with MLC retracted. All
measurements were done for SSD = 100 cm. Information on the differences in the construc-
tion of the linac head components have been described in table 1 and in Martínez (2013).

Discussion

The agreement between the simulated and measured data for depth-dose profile is very good,
with a difference below 5%. Secondaries contribute about 40% to the surface dose. At the
depth of maximum dose, their contribution decreases to 10% and then slowly decreases to a
value of about 8% at the depth of 50 cm.
The lateral profiles obtained with PHITS agree within a few percent infield with the meas-
ured data by Kaderka et al (2012) and Mesbahi (2006). For out-of-field the calculated data
mostly tended to overestimate the dose, ranging between  ±30% with the average absolute
difference between measured and calculated dose of 11% and 5% in-plane and of 10% and
12% cross-plane at 3 cm and 10 cm depths respectively. The average absolute difference

N266
Phys. Med. Biol. 60 (2015) N261 Note

50000
Surface
isocenter
30 cm from isocenter
40000
10 cm depth
isocenter
Fluence (n/cm /MU/u) 30 cm from isocenter
30000
2

20000

10000

0
0.01 0.1 1 10
Energy (MeV)

Figure 4. PHITS simulated neutron fluences for an isocenter and for a location 30 cm
from the isocenter at the water surface and at the depth of 10 cm for ELEKTA SL25.

between the calculated and measured photon dose for the out-of-field region for both planes
was around 11%.
Irrespective of the water depth and distance to the isocenter, all neutron spectra have the
same profile, with a peak at about 0.5–1 MeV. Compared to the isocenter, the total number of
neutrons scored at a location 30 cm out-of-field is about 40% lower at the water surface and at
10 cm depth. After traversing 10 cm in water the total neutron fluence is attenuated by around
70% compared to the surface, for both locations, the isocenter and 30 cm from the isocenter.
The neutron spectra are characterized by an evaporation component, with a peak in the
range 200–700 keV, and a component from the relatively weak direct reactions in the several
MeV energy range (Zanini et al 2004). Here the direct component has a peak at about 2 MeV
and overlaps with the evaporation component producing a complex peak at about 1 MeV.
Since the high-energy neutrons (direct component) are produced in forward directions, their
number registered out-of-field at the surface of the water phantom was reduced compared to
the number of high-energy neutrons registered infield region. Furthermore, it was observed
that after passing 10 cm in water, a significant decrease of both evaporation and direct compo-
nents occurred as a consequence of neutron thermalization in matter.
The agreement between PHITS neutron spectra simulations and measurements by Kaderka
et al (2012) is rather poor. It could be explained by the rather high uncertainty of the BDS
measurements; in the order of 20% and even larger in the low energy region (Kaderka et al
2012). The statistical uncertainties of PHITS simulations are within 6 and 20%, depending on
the energy range. Despite the differences in the field size and linac models and manufactur-
ers, the PHITS data agreement with neutron profiles measured by Ongaro et al (2000) and
Zanini et al (2004) is very good. After closing jaws and MLC, Howell et al (2009) reported a
total reduction of the direct component of neutron characteristics at 40 cm from the isocenter.
Retracted MLC, as measured by Kry et al (2007), gives increase of the production of lower
energy neutrons with energies between 10 keV and 500 keV.

N267
Phys. Med. Biol. 60 (2015) N261 Note

5
1x10
Ongaro et al. 2000 (8 cm) surface, 8 - 15 cm
4 Zanini et al. 2004 (15 cm) from the isocenter
8x10
4x10
4
Kaderka et al. 2012 (10.5 cm)
PHITS (10 cm)
3x10
4 (a)

4
2x10
Fluence (n/cm /MU/u)

4
1x10
2

0
4
8.5x10
surface, 30 - 40 cm
Kry et al. 2007 (40 cm)
8.0x10
4 from the isocenter
Howell et al. 2009 (40 cm)
3.0x10
4
Kaderka et al. 2012 (30.5 cm)
PHITS (30 cm)

2.0x10
4 (b)

4
1.0x10

0.0
0.01 0.1 1 10
Energy (MeV)

Figure 5. Neutron energy spectrum simulated by PHITS and measured at the surface of
the water phantom between 8–15 cm (a) and 28–40 cm from the isocenter (b).

Conclusions

The MC code PHITS has been tested to calculate the dose distributions and neutron spectra
for ELEKTA SL25 medical linear accelerator, and the results were compared to measured
data. The calculations presented in this report show that PHITS is a reliable tool for calculat-
ing the dose profiles and energy spectra for photon beams. The calculated photon doses for
out-of-field region were, on average, 11% different from the measured doses. We consider the
photon agreement of our model at 18 MV to be quite good and well suited to calculate the
out-of-field doses.
The neutron spectra confirm that most neutrons are produced at energies between 100 keV
and 1 MeV, where the biological effectiveness is high, as pointed out in the publication 103 of
the International Commission on Radiological Protection (ICRP 2007).
After analyzing the PHITS calculations and the experimental data by Kaderka et al (2012),
one can observe that (i) the measured spectra are extremely narrow with peak around 1 MeV,
(ii) the evaporation component of the neutron spectrum is very small, and (iii) neither agree
with simulations presented here nor with other measurements found in the literature. Since
there is only limited published data for ELEKTA machines, it could be recommended to repeat
the measurements at KGU.

N268
 Phys. Med. Biol. 60 (2015) N261

Table 2. Comparison of the calculated and measured integral neutron fluence obtained in this work and in the literature.

Distance to Calculated
Authors Machine Field size (cm2) isocenter (cm) Measured (107 n cm−2 Gy−1) (107 n cm−2Gy−1)
This work ELEKTA SL25 18 MV 5   ×   5 0 — 0.97
10 — 0.65
30 — 0.56

N269
Howell et al (2009) ELEKTA Precise 10   ×   10 0 1.58 —
18 MV Jaws closed 40 0.40 —
Kry et al (2007) Varian 2100 18 MV 3.6   ×   4 MLC retracted 30 0.81 1.09
Kaderka et al (2012) ELEKTA SL25 18 MV 5   ×   5 0 1.24 —
10.5 0.45 —
30.5 0.35 —
Note
 Phys. Med. Biol. 60 (2015) N261 Note

Conflict of interest

The authors report no conflicts of interest. The authors alone are responsible for the content
and writing of the paper.

Acknowledgments

This work was partly funded by the European project ALLEGRO (grant agreement no. 231965).
The authors wish to express gratitude to D Schardt and R Kaderka (GSI) for their support dur-
ing the ELEKTA SL25 model development. The simulations were performed on resources pro-
vided by the Swedish National Infrastructure for Computing (SNIC) at Linköping University.

References

Cullen D E, Hubbell J H and Kissel L 1997 EPDL97: the evaluated photon data library ‘97 version
UCRL-50400 Lawrence Livermore National Laboratory, vol 6, Rev. 5 (https://www-nds.iaea.org/
epdl97)
Fuller E G 1985 Photonuclear reaction cross sections for 12C, 14N and 16O Phys. Rep. 127 185–231
Halbleib J A, Kensek R P, Mehlhorn T A and Valdez S M 1992 ITS Version 3.0: The Integrated TIGER
Series of Coupled Electron/photon Monte Carlo Transport Codes Sandia National Laboratories
SAND91-1634
Howell R M, Kry S F, Burgett C F, Hertel N E and Followill D 2009 Secondary neutron spectra from
modern Varian, Siemens and Elekta linacs with multileaf collimators Med. Phys. 36 4017–38
ICRP 2007 The 2007 Recommendations of the International Commission on Radiological Protection.
ICRP Publication 103 Ann. ICRP 37 (2–4)
Kaderka R 2011 Out-of-field dose measurements in radiotherapy PhD Thesis
Kaderka R et al 2012 Out-of-field dose measurements in a water phantom using different radiotherapy
modalities Phys. Med. Biol. 57 5059–74
Kry S F, Titt U and Followill D 2007 A Monte Carlo model for out-of-field dose calculation from high-
energy photon therapy Med. Phys. 34 3489–99
Martínez Ovalle S A 2013 Neutron dose equivalent in tissue due to linacs of clinical use Frontiers in
Radiation Oncology ed T Kataria (chapter 6)
Mesbahi A 2006 Development a simple point source model for Elekta SL-25 linear accelerator using
MCNP4C Monte Carlo code, Iran J. Radiat. Res. 4 7–14
Nuclear Data Center (JAEA) 2013 JENDL-4.0u (JENDL-4.0 Updated Files) JENFL-4.0+ (JENDL-4.0
Plus Files) (Fukui: Japan Atomic Energy Agency) (www.ndc.jaea.go.jp/jendl/j40/update/)
Ongaro C, Zanini D A, Nastasi S U, Rodenas S J, Ottaviano G and Manfredotti C 2000 Analysis of
photoneutron spectra produced in medical accelerators Phys. Med. Biol. 45 55–61
Perkins S T, Cullen D E, Seltzer S M 1991 Tables and graphs of electron-interaction cross sections from
10 eV to 100 GeV derived from the LLNL evaluated electron data library (EEDL), Z = 1–100
Lawrence Livermore National Laboratory, UCRL-50400, p 31
Puchalska M, Sihver L, Sato T, Berger T and Reitz G 2012 Simulations of MATROSHKA experiment at
ISS using PHITS Adv. Space Res. 50 489–95
Sato T, Kase Y, Watanabe R, Niita K and Sihver L 2009 Biological dose estimation for heavy ion therapy using
an improved PHITS code coupled with the microdosimetric kinetic model Radiat. Res. 171 107–17
Sato T et al 2013 Particle and heavy ion transport code system PHITS, version 2.52 J. Nucl. Sci. Technol.
50 913–23
Swanson W P 1978 Calculation of neutron yields released by electrons incident on selected materials
Health Phys. 35 353–67
Takam R, Bezak E, Marcu L G and Yeoh E 2011 Out-of-field neutron and leakage photon exposures and
the associated risk of second cancers in high-energy photon radiotherapy: current status Radiat.
Res. 176 508–20
Zanini A et al 2004 Monte Carlo simulation of the photoneutron field in linac radiotherapy treatment
with different collimation systems Phys. Med. Biol. 49 571–82

N270

View publication stats