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

net/publication/229308261

Estimation of neutron production from accelerator head assembly of 15MV

medical LINAC using FLUKA simulations

Article in Nuclear Instruments and Methods in Physics Research Section B Beam Interactions with Materials and Atoms · December 2011
DOI: 10.1016/j.nimb.2011.04.013

CITATIONS READS

20 2,513

6 authors, including:

Bhushankumar J Patil Sharad T Chavan

Abasaheb Garware College, Savitribai Phule Pune University, Pune Society for Applied Microwave Electronics Engineering & Research
48 PUBLICATIONS 294 CITATIONS 31 PUBLICATIONS 140 CITATIONS

SEE PROFILE SEE PROFILE

S.N. Pethe Vasant Bhoraskar

Society for Applied Microwave Electronics Engineering & Research Savitribai Phule Pune University
25 PUBLICATIONS 116 CITATIONS 256 PUBLICATIONS 2,352 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Sharad T Chavan on 01 July 2019.

The user has requested enhancement of the downloaded file.

This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
 Author's personal copy

Nuclear Instruments and Methods in Physics Research B 269 (2011) 3261–3265

Contents lists available at ScienceDirect

Nuclear Instruments and Methods in Physics Research B

journal homepage: www.elsevier.com/locate/nimb

Estimation of neutron production from accelerator head assembly of 15 MV

medical LINAC using FLUKA simulations
B.J. Patil a, S.T. Chavan b, S.N. Pethe b, R. Krishnan b, V.N. Bhoraskar a, S.D. Dhole a,⇑
a
Department of Physics, University of Pune, Pune 411 007, India
b
SAMEER, IIT Powai Campus, Mumbai 400 076, India

a r t i c l e i n f o a b s t r a c t

Article history: For the production of a clinical 15 MeV photon beam, the design of accelerator head assembly has been
Available online 20 April 2011 optimized using Monte Carlo based FLUKA code. The accelerator head assembly consists of e–c target,
flattening filter, primary collimator and an adjustable rectangular secondary collimator. The accelerators
Keywords: used for radiation therapy generate continuous energy gamma rays called Bremsstrahlung (BR) by
LINAC impinging high energy electrons on high Z materials. The electron accelerators operating above
FLUKA 10 MeV can result in the production of neutrons, mainly due to photo nuclear reaction (c, n) induced
Electron
by high energy photons in the accelerator head materials. These neutrons contaminate the therapeutic
Bremsstrahlung
Neutron
beam and give a non-negligible contribution to patient dose. The gamma dose and neutron dose equiv-
Accelerator head assembly alent at the patient plane (SSD = 100 cm) were obtained at different field sizes of 0  0, 10  10, 20  20,
30  30 and 40  40 cm2, respectively. The maximum neutron dose equivalent is observed near the
central axis of 30  30 cm2 field size. This is 0.71% of the central axis photon dose rate of 0.34 Gy/min
at 1 lA electron beam current.
Ó 2011 Elsevier B.V. All rights reserved.

1. Introduction effect occurring is the production of neutrons, mainly due to the

photonuclear giant-dipole-resonance (GDR) reaction (c, n) induced
The MeV energy Bremsstrahlung produced by medical acceler- by high energy photons in the accelerator head materials [1]. For
ators is a common form of treatment modality for malignant conventional treatment techniques, the contamination is neglected
tumors that occur at depth below the skin surface. The linear for the patient and only accounted for radiation protection. How-
accelerator (linac) is a primary tool in external beam radiotherapy, ever, if precision radiation treatments like intensity modulated
which generates continuous energy gamma rays called radiation therapy (IMRT) are used, then the leakage and neutron
Bremsstrahlung (BR) by impinging electrons on high Z material radiation increases, as these techniques require longer beam-on
(e–c target). The clinically applicable photon beam is produced times. It is predicted in the literature that additional dose due to
in an e–c target, flattened with a flattening filter, collimates in pri- the photoneutrons is proportional to the beam-on time [2]. The
mary collimator and beam shaping using secondary collimator. The biological effectiveness of neutrons is substantially higher than
main challenge while using such clinical photon beam in tumor that of photons [3], therefore even a small neutron dose will
treatment is the application of high doses to the tumorous body increase the risk for secondary cancer. Therefore, it is necessary
regions by simultaneous sparing of the healthy tissues. Dual trans- to minimize the contribution of neutrons while designing such
mission ionization chambers are used for monitoring the photon accelerator head assembly. In addition, knowledge of the energy
radiation beam output as well as the radial and transverse beam spectrum of the photo neutron contamination allows one to esti-
flatness. Also, this helps to get an exact idea of dose delivered to mate the equivalent neutron dose received by the patient, optimize
the patient. the room shielding and even the use of better energy-dependent
Photon beams with energies higher than 10 MeV are preferred, quality factors to estimate neutron doses received by the medical
if doses should be delivered to larger depths (e.g. for the treatment personnel working in and around therapy facilities.
of prostate cancer) and to enhance the skin sparing. But a parasitic Neutron leakage from radiotherapy accelerators has been inves-
tigated as early as in 1951 [4]. The guidelines have been recom-
mended regarding maximum admissible contamination levels by
⇑ Corresponding author. Tel.: +91 20 25692678; fax: +91 20 25691684.
the NCRP [5]. The International Electrochemical Commission
E-mail addresses: bjp@physics.unipune.ac.in (B.J. Patil), sharad@sameer.gov.in
(S.T. Chavan), sanjay@sameer.gov.in (S.N. Pethe), krishnan@sameer.gov.in (R.
(IEC) recommends certain limits for neutron absorbed dose in the
Krishnan), vnb@physics.unipune.ac.in (V.N. Bhoraskar), sanjay@physics.unipune.ac. patient plan [6], and the American Association of Physicists in
in (S.D. Dhole). Medicine (AAPM) has reported a review on neutron measurement

0168-583X/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.nimb.2011.04.013
 Author's personal copy

3262 B.J. Patil et al. / Nuclear Instruments and Methods in Physics Research B 269 (2011) 3261–3265

methodologies [7]. Fast neutron contamination of high-energy

Bremsstrahlung beams in radiotherapy has been investigated by
McGinley et al. [8]. Price et al. has calculated and measured the
neutron dose equivalent per gamma dose using the different med-
ical linac facilities at different electron energies [1].
In the present paper, an objective was to design the accelerator
head assembly consisting of e–c target, primary collimator,
secondary collimator (X and Y jaws) and flattening filter for
15 MeV medical linac. In addition, the neutron contamination in
photon beam has been estimated in terms of dose equivalent and
energy spectra. For this work, FLUKA simulations has been carried
out to evaluate the photoneutron yield and spectra produced
through the accelerator head assembly of 15 MeV medical linac
as a function of the radiation field sizes.

2. Materials and methods

NCRP Report No. 79 addressed the spectrum of neutrons pro-

duced by a Bremsstrahlung radiation with a maximum energy up
to the incident electron energy [9]. The physics model used in this
study considered only the production of neutrons. But, in the
whole accelerator facility, neutrons are transported through the
target and moderated or absorbed by light nuclei. These complex
processes are difficult to study theoretically, even on the basis of
correct experiments. Also, measurements of a neutron spectrum
produced in the linear accelerator based neutron source is difficult
to perform with the standard nuclear instrumentation, due to the
high fluence rate of photons with respect to neutrons and the
pulsed radiation field. Therefore a computer code allowing a suit-
able simulation of the entire process of photoneutron generation
and transport across the accelerator head represents a useful tool Fig. 1. Schematic of various components of the proposed accelerator head assembly
to evaluate the undesired neutron leakage dose at the patient. As (not to the scale).

FLUKA can handled the accurate electron–nucleus, electron–

electron Bremsstrahlung and photo nuclear interactions (described
maximum field of 51 cm diameter at patient plane, the primary
by Vector Meson Dominance, Delta Resonance, Quasi-Deuteron
conical collimator comprises a 28° cone bored in a metal block.
and Giant Dipole Resonance model) over the whole energy range
The thickness of the shielding block is usually designed to
[10–13] is the best choice for simulation. Monte Carlo simulations
attenuate the mean primary X ray beam intensity to be less than
with FLUKA were carried out to estimate undesired neutron
0.1% of the initial value (three tenth-value layers (TVLs)). According
leakage dose at the patient due to various accelerator head
to IEC’s recommendations, the maximum leakage should not
components.
exceed 0.2% of the open beam value. A more uniform angular dis-
The linear accelerator (LINAC) [14] propels electron along
tribution of the photon beam can be achieved by passing it through
straight line trajectories by the applications of alternating electric
a flattening filter. The flattening filter is like of Gaussian shaped.
fields. The microwave field having frequency either 1.5 or 3 GHz,
The maximum circular field defined by primary collimator are
is used to propagate the beam. Electrons after passing through
truncated with an adjustable rectangular collimator which consists
microwave accelerating cavities, gains energy and coming out with
of upper and lower independent movable jaws for producing
beam energy of 6 MeV. The parameters of the electron beam are
rectangular and square fields with a maximum dimension of
pulsed current 130 mA, pulse width 4.5 ls and pulse rate 150–
40  40 cm2 at the linac iso-centre. These blocks have sufficient
200 PPS. Therefore, the average current of the electron beam is
thickness to shield out unwanted radiations. The upper jaws move
100 lA.
in the in-plane direction and the lower jaws move in cross-plane
direction. The IEC recommends that the transmission of the
3. Proposed accelerator head assembly primary photon beam through the rectangular collimator should
not exceed 2% of the open beam value. The material and dimension
Clinical photon beams emanating from a medical linac are pro- of collimators were optimized such that the neutron contamina-
duced in the e–c target flattened with a flattening filter and colli- tion in the gamma beam was below the allowed limit.
mates in collimators. The photon collimation and beam shaping is
achieved with primary collimator and movable secondary collima-
tor. The various components of accelerator head assembly and the 4. Results and discussion
beam shaping using primary and secondary collimator is shown in
Fig. 1. The Bremsstrahlung yield produced in e–c target depends on Firstly, to optimize the e–c target Bremsstrahlung fluence from
the electron energy and the atomic number of the target element. various high Z target has been calculated. Amongst the materials
Therefore, high atomic number elements are used as e–c target studied, tungsten was found to be the best suitable as e–c target
and immediately after that there is primary collimator. It is de- because of its physical properties like melting point, heat conduc-
signed to absorb all unwanted sections of the X-ray field. The cir- tivity and highest Bremsstrahlung yield. The 0.42 cm thick
cular conical hole defines the maximum divergence of the beam tungsten has been optimized as an e–c target for 15 MV medical
and therefore the maximum circular field size. For obtaining LINAC since it absorbs almost all the incident electrons. The
 Author's personal copy

B.J. Patil et al. / Nuclear Instruments and Methods in Physics Research B 269 (2011) 3261–3265 3263

at 100 cm from the source to surface distance (SSD). The increase

in thickness of primary collimator decreases the penumbra. In
addition, the leakage radiations were calculated offside at a dis-
tance of 1 m from the beam centre and found to be less than
0.2% of beam value (recommended by IEC) for the thickness more
than 8 cm of primary collimator. Therefore, it was optimized to use
10 cm thickness of W–Cu for primary collimator to minimize the
penumbra. The neutron fluence calculated at iso-centre is
3.94  10 9 neutron cm 2/e. The unflattened absorbed dose distri-
bution can be modelled as Gaussian along a plane transversal to
the beam axis. Therefore, a Gaussian-shaped filter may reflect a
smoothly increased attenuation towards the central beam axis. A
Gaussian shaped filter was divided in eight truncated right angle
cone (TRC) as shown in Fig. 4(a). For the different values of height,
base radius and top radius of each TRC, the FLUKA simulations
were carried out to obtain flattened absorbed dose in water phan-
tom at SSD = 100 cm. The beam profile was a key parameter for the
design of a flattening filter. As seen from Fig. 3, less number of
neutrons are produced from iron material as compared to lead
Fig. 2. Bremsstrahlung spectra at e–c target and collimator surface for 15 MeV material, therefore, iron has been used as a filter. The dimensions
incident electrons. of optimized flattening filter made of eight TRC’s are shown in
Fig. 4(a). The 3D drawing of the optimized iron flattering filter is
shown in Fig. 4(b) which is plotted in Simplegeo 4.2 [15] and it
Bremsstrahlung spectrum estimated on the target surface and the gives flattened dose for 40  40 cm2 field size.
collimator incident face is shown in Fig. 2. It is observed from the Jaws of the secondary collimator were positioned such that the
figure that the Bremsstrahlung spectra have peak at 0.5 MeV en- rotation of respective X-jaws and Y-jaws forms square field size.
ergy and have continuous energy spectrum up to the incident elec- The thickness of secondary collimator was optimized such that
tron energy. The mean energy of the Bremsstrahlung spectrum is
around 2.033 MeV. The FLUKA simulations were performed to find
out the tenth value layer (TVL) thickness in various materials for
the 15 MeV electron beam generated Bremsstrahlung spectrum.
From the simulation, the TVL values for iron, lead, tungsten, bis-
muth, tungsten + copper, tantalum have been estimated and they
are 7.64, 3.87, 2.93, 4.40, 3.37, 3.30 cm, respectively. The neutron
fluence produced through photonuclear reaction in these materials
is given in Fig. 3. The material having less TVL thickness and low
neutron production is the best material to be used for primary col-
limator. Therefore, the W–Cu material has been optimized to de-
sign the primary collimator. To attenuate the Bremsstrahlung
beam intensity to less than 0.1% of initial value, W–Cu thickness
has been optimized to more than three TVL thicknesses.
The simulations were carried out using a block of W–Cu having
conical opening of 28°, for the calculation of Bremsstrahlung radi-
ations at iso-centre and leakage radiation. The iso-center is defined

Fig. 4. (a) Gaussian shape flattening filter (b) 3D drawing of flattening filter for
Fig. 3. Variation in neutron fluence generated as a function of material thickness. 15 MeV linac.
 Author's personal copy

3264 B.J. Patil et al. / Nuclear Instruments and Methods in Physics Research B 269 (2011) 3261–3265

Fig. 5. Schematic of the accelerator head assembly modelled in FLUKA for 15 MV medical LINAC (not to the scale).

Table 1
Bremsstrahlung fluence, Bremsstrahlung dose delivered in water and neuron fluence
for the different filed sizes.

Field Bremsstrahlung fluence Bremsstrahlung dose Neutron fluence

size (photon cm 2 /e ) (Gy min 1 lA 1) (neutron cm 2/e )
(cm2)
5 9
10  10 7.216  10 0.293 2.452  10
5 9
20  20 8.051  10 0.323 3.408  10
5 9
30  30 8.652  10 0.342 3.612  10
5 9
40  40 8.974  10 0.335 3.518  10

Fig. 6. Relative photon depth dose distribution for the various field sizes at SSD of
100 cm.

Fig. 8. The ratio of neutron dose equivalent to central axis photon absorbed dose
estimated at the patient plane for different field sizes.

the transmission of primary X-ray beam should not exceed 2% of

the open beam value. Therefore, thickness of the secondary colli-
mator was optimized to 8 cm. The optimized value of e–c target,
primary collimator, filter and secondary collimator, the structure
was modelled in FLUKA as shown in Fig. 5. Using the trigonometry,
position of the X and Y jaws has been calculated for different field
sizes. The rotation of X and Y jaws of secondary collimator along
Fig. 7. Flattened dose profile in water phantom for the various field sizes. the arc changes the radiation field size area from 0  0 to
 Author's personal copy

B.J. Patil et al. / Nuclear Instruments and Methods in Physics Research B 269 (2011) 3261–3265 3265

dose equivalent estimated are consistent with the results of other

measurements reported in literature [16] and fall within the
allowed limit by IEC. In addition, the ratio of neutron dose equiva-
lent to the central axis photon dose was maintained below the
allowed limit set by IEC (<1 mSv/Gy) inside and 0.5 mSv/Gy outside
of photon field. The neutron fluence spectrum calculated for differ-
ent field sizes using FLUKA is shown in Fig. 9.

5. Conclusion

For the production of 15 MeV photon beam for clinical applica-

tions, the design of an accelerator head assembly has been
proposed and optimized. Using optimized design, the flattened
dose calculated at 100 cm SSD is 0.34 Gy/min at 1 lA electron
beam current for 30  30 cm2 field size. The maximum square field
size can be produced by the collimator is 30  30 cm2. In addition,
the neutrons produced in the accelerator head assembly and the
ratio of the neutron dose equivalent to the central axis photon dose
have been estimated. The ratio was found to be below the allowed
Fig. 9. The neutron energy spectra at iso-center for different field sizes. limit recommended by IEC is <1 mSv/Gy.
40  40 cm2. Using calculated positions of the jaws for each field
size, the accelerator head assembly has been modeled in FLUKA References
to estimate the gamma absorbed dose and neutron dose equiva- [1] K.W. Price, R. Nath, G.R. Holeman, Med. Phys. 5 (4) (1978) 285–289.
lent. In addition, the neutron fluence and respective spectra for dif- [2] R.M. Howell, M.S. Ferenci, N.E. Hertel, G.D. Fullerton, Med. Phys. 32 (3) (2005)
ferent field sizes have also been estimated. Fig. 6 shows the relative 779–786.
[3] National Council on Radiation Protection and Measurement (NCRP), Report No.
photon depth dose distribution for 10  10, 20  20, 30  30,
116, Washington, DC, 1993.
40  40 cm2 field sizes at SSD of 100 cm using the optimized accel- [4] J.S. Laughlin, Nucleonics 8 (1951) 5.
erator head geometry. The distance at which maximum dose deliv- [5] National Council on Radiation Protection and Measurements (NCRP), Report
79, Washington, DC, 1984.
ered in water is 2 cm. The Fig. 7 shows dose profiles in water
[6] International Electrotechnical Commission. International Standard IEC 60601–
phantom for different field sizes. 2–1, 1998.
Table 1 shows the Bremsstrahlung fluence, maximum Brems- [7] American Association of Physicists in Medicine, Report 19, American Institute
strahlung dose delivered in water and neuron fluence for different of Physics, New York, 1986.
[8] P.H. McGinley, J.C. Landy, Phys. Med. Biol. 34 (1989) 777–783.
field sizes. It is observed from the results that maximum the field [9] NCRP report No. 79, Neutron contamination from medical electron accelerator,
size more the dose delivered. However, in case of 40  40 cm2 field 1984.
size the dose obtained to be reduced as compared to 30  30 cm2 [10] A. Fasso, A. Ferrari, J. Ranft, P.R. Sala, OECD Documents (1995) 287–304.
[11] A. Fasso, A. Ferrari, P.R. Sala, CERN internal report TIS-RP/IR/95-27, (1995)
field size because of the effective area offered by conical beam 643–649.
which is less than the square field size. It is also observed that [12] A. Fasso, A. Ferrari, P.R. Sala, OECD-NEA (1998) 61–74.
the neutron fluence increases with field size. The ratio of neutron [13] A. Fasso, A. Ferrari, P.R. Sala, AIP conf. Proc. 769 (2005) 1303–1306.
[14] R. Krishnan, A.P. Deshpande, T.S. Dixit, S. Chavan, C.S. Nainwad, S.N. Pethe, T.T.
dose equivalent to central axis photon absorbed dose at the patient Tiwari, Proc. PAC09, Vancouver, BC, Canada, 2007.
plane for different field sizes is shown in Fig. 8. The maximum neu- [15] C. Theis, K.H. Buchegger, M. Brugger, D. Forkel-Wirth, S. Roesler, H. Vincke,
tron dose equivalent is observed near the central axis of Nucl. Instr. Meth. A 562 (2006) 827–829.
[16] N. Golnik, M. Zielczynski, W. Bulski, P. Tulik, T. Pałko, Rad. Prot. Dos. 126 (1–4)
30  30 cm2 field size. This is 0.71% of the central axis photon dose
(2007) 619–622.
rate of 0.34 Gy/min at 1 lA beam current. The values of neutron

View publication stats