Dosimetric characteristics of 6 MV photons from TrueBeam STx medical linear accelerator: Simulation and experimental data

Nuclear Science and Technology, Vol.9, No. 2 (2019), pp. 37-44

Dosimetric characteristics of 6 MV photons from TrueBeam STx

medical linear accelerator: simulation and experimental data

N. D. Ton¹, B. D. Linh^1,*, Q.T. Pham²

¹Institute for Nuclear Science and Technology, 179 Hoang Quoc Viet, Cau Giay, Ha Noi

²Department of Radiation Oncology and Radiosurgery, 108 Military Central Hospital, Ha Noi

^*Corresponding author, Email: [email protected]

(Received 17 October 2019, accepted 14 November 2019)

Abstract: A TrueBeam STx is one of the most technologically advanced linear accelerators for

radiotherapy and radiosurgery. The Monte Carlo simulation widely used in many applications in

various fields such as nuclear physics, astrophysics, particle physics, and medicine. The

Geant4/GATE Monte Carlo toolkit is developed for the simulation in imaging diagnostics, nuclear

medicine, radiotherapy, and radiation biology to more accurately predict beam radiation dosimetry. In

this work, we present the simulation results of the dosimetric characteristics of a 6 MV photon beam

of TrueBeam STx medical LINAC using Monte Carlo Geant4/GATE. The percentage depth dose

(PDD), central axis depth dose (Profile) have been simulated and compared with those measured in a

water phantom for field sizes 10×10 cm²via the gamma-index method. These results will permit to

check calculation data given by the treatment planning system.

Keywords: Geant4/GATE, Radiotherapy, TrueBeam STx, Phase Space file, PDD, Profile.

anatomical sites, where the electron transport

approximations from analytic or semi-analytic

dose calculation algorithms are not accurate

enough. The long computation time becomes

a reason to hinder MC simulation from

widespread use in the routine clinical practice

[2-4]. However, due to industrial secrets, this

information is sometimes unavailable to the

I. INTRODUCTION

The characteristics of the photon beam

as Percentage Depth Dose and Profile in a

medical linear accelerator (Linac) are essential

for building the beam model and guarantee the

dose distribution for input into Treatment

Planning System (TPS) [1]. TrueBeam STx is

one of the latest generation Linac of Varian It

has different characteristics with the previous

Linac such as two modes of photon Flattening

Filter (FF) and Flattening Filter Free (FFF).

TrueBeam STx can be used for many forms of

advanced treatment techniques including

image-guided radiotherapy (IGRT), intensity-

modulated radiotherapy (IMRT) and RapidArc

radiotherapy technology.

general

medical

physics

community.

TrueBeam STx Linac of Varian Medical

Systems is an example. Instead, Phase Space

files (PhS) for each of the beam have been

made available to the Varian TrueBeam STx

users [5]. Although the accuracy of these PhS

among different TrueBeam STx installations

was proven [6-8], it is advisable to validate

them by comparing the dose distributions

produce in the MC simulation with

experimental measurements performed on a

TrueBeam STx Linac. Besides, the

verification of the geometry in this machine is

also required.

Monte Carlo (MC) method is a

popular method to estimate accurate dose

distributions

for

clinical

beams

in

radiotherapy. Especially, the method is

important for planning heterogeneous

DOSIMETRIC CHARACTERISTICS OF 6 MV PHOTONS FROM TRUEBEAM STX …

Geant4/GATE MC code is one of the

most popular MC open-source applied

nowadays in medical physics, especially the

radiotherapy field [9, 10]. Much scientific

research has been carried out to accurately

simulate complex geometry in order to find an

imaging and Cone Beam computed

tomography (CBCT) capability. One of the key

features is the availability of two modes of

photon beams: FF (Flattening Filter) and FFF

(Flattening Filter Free). FFF beams delivered

with ―conventional‖ medical linear accelerators

have the conical flattening filter removed and

replaced by a thin copper foil (about 1mm

thick). This foil is the same for all energies.

Truebeam STx having 6, 8, 10, 15 MV

standard FF and 6 MV, 10 MV with the new

FFF beam for photon beams. FFF beams offer

a maximum high dose rate of 1400 MU/min for

6 MV FFF and 2400 MU/min for 10 MV FFF,

respectively [15-17].

overall reasonable

agreement

between

calculated and experimental doses for Linac

[11-14]. In Vietnam, this study is the first one

to assess the capabilities of Geant4/GATE in

the prediction of characteristics of the

TrueBeam STx Linac 6 MV photon beam.

The aim of this work is to evaluate the

accuracy of Geant4/GATE in simulating

characteristics of

6

MV photons from

TrueBeam STx Linac. In order to establish this

goal, PDD and Profile were investigated by

Geant4/GATE with Varian PhS and were

measured by a Blue Phantom system (The Blue

Phantom 2, CC013 ionization chamber,

RAZOR chamber, and electronic equipment).

Simulations and experimental results were

compared in specifications of dose difference

relative to the maximum dose of the

measurement, we explain the details of this in

section III. The gamma index was also

performed using a 2%/2mm standard of dose-

difference (DD) and distance-to-agreement

(DTA).

B. The Phase Space files

The TrueBeam STx Phase Space files

for 6 MV FF with the International Atomic

Energy Agency (IAEA) format [18] have been

made by Varian using the Geant4 MC toolkit

[19]. These files were verified as radiation

sources to allow users to perform accurate

simulations, which were recorded immediately

upstream of the movable at a distance of 73.3

cm from the isocenter. Therefore, users are

required to code into their MC application only

the geometrical details of the components of

the treatment head below the PhS surface,

including the jaws and MLC. Each PhS file of

6 MV FF beam contains 10⁹original histories

and 5×10⁷particles. The initial source is an

electron beam. The energy spectrum of this

source is non-Gaussian with a peak at 6.13

MeV and a tail extending up to 6.35 MeV. In

order to speed up the accelerator head

simulation, PhS file stored three types of

particles including photon, electron, positron.

The details of the header of each file are

reported [5]. The simulation has been used not

only photon but also electron and positron in

PhS. This allows considering electron and

positron contaminations from the photon beam.

II. MATERIALS AND METHODS

A. The Varian TrueBeam STx

TrueBeam STx with High Definition

120-MLC (Multi-Leaf Collimator) has at

central 32 leaf pairs of 2.5 mm and 28 leaf

pairs of 5 mm leaf thickness to achieve high

precise target conformation and minimize the

penumbra effect, which can be used for

Stereotactic Radiosurgery (SRS), Stereotactic

Radiation Therapy (SRT) and Stereotactic

Body Radiation Therapy (SBRT). The

TrueBeam STx is also equipped with kV/MV

38

N. D. TON, B. D. LINH, Q.T. PHAM

maximum dose (1.5 cm) scans were made for

C. Experimental measurements

FF 6 MV with 10×10 cm²field size. To avoid

any ripple effect in the measurement, a PDD

scan was started from the bottom of the tank

moving toward the surface of the water. Data

processing and analysis were performed using

IBA’s OmniPro Accept with application

setting for a geometric mean smoothing

function with a value of 3 mm. Appropriate

stopping power ratio factors were used for

electron ionization values to PDD values

conversion. Measured PDD curve was

compared to Golden Beam Data (GBD) and

the reproducibility of the measurement data.

GBD was provided by Varian, which often use

The Blue Phantom 2 in a large

48×48×48 cm³water phantom a three

dimensions (3D) scanning system was used

[20]. Fig. 1 was shown actually a photo of Blue

Phantom 2 under TrueBeam STx head in the

experiment. The scan in water was acquired at

the same source-to-surface distance (SSD) =

100 cm. Beam scanning and collecting data

were performed in accordance with

professional guidelines, such as AAPM Task

Group [21, 22]. The ionization chamber CC13

and RAZOR chamber [20] were used for beam

data collection and dosimetric measurement

following reference [22] recommendation.

CC13 is a cylindrical ionization chamber with

a sensitive volume of about 0.13 cm³and an

inner diameter of the outer electrode about 6.0

mm. The effective point offset in the

measurement of this chamber is 1.8 mm for 6,

8 MV and 2 mm for 10, 15 MV photon beams.

The RAZOR is a compact chamber for

measurements of small fields and of ranges

with high dose gradients. The RAZOR has

sensitive volume and the radius cavity are 0.01

cm³and 1.0 mm, respectively.

for

commissioning

measurements.

A

comparison of the PDD with the Gamma Index

criteria of 2%/2 mm yielded a gamma pass rate

of 97%. Therefore, the accuracy of this

measurement can be considered to be within

2%/2 mm.

D. Geant4/GATE

For many years, the Geant4-based

GATE MC code has been developed as an

open-source MC program for nuclear medicine

simulation, with a focus on PET and SPECT

imaging [23]. This toolkit allows creating a

simulation on the basis of simple macro-

command instead of handling tedious C++

syntaxes of Geant4 code. It helps a quicker

learning phase for new users and makes a small

size of the GATE work folder easy to share

within the community. Details about the GATE

capabilities and validation are presented

elsewhere [23-25].

A new GATE v8.2 was used for this

simulation [26]. 6 Varian PhS files of smaller

size (2 GBs) were imported into GATE and

used for the downstream of the jaws as a

source in the Linac. These individual files were

then concatenated to one large PhS file. After

exiting the PhS plane, the particle passes

Fig.1. Photograph of Blue Phantom (IBA,

Germany) scanning water phantom below

TrueBeam STx head in the experimental setup.

Measurements include PDD along the

central axis and crossline profile at the depth of

39

DOSIMETRIC CHARACTERISTICS OF 6 MV PHOTONS FROM TRUEBEAM STX …

through the second collimator as the Y and X

jaws and MLC. Data for the material and

geometry of the Linac components were

obtained from the TrueBeam STx Monte Carlo

package [19].

final assessment of the dose distribution

quality. For the regions of significant

disagreement, the Gamma Index value is

greater than unity that will be apparent relative.

The gamma pass rate was defined as a

quotient of the passing points and all points.

For the global Gamma Index passing criteria of

2%/2 mm, a good agreement, a high

Geant4 Electromagnetic physics package

3 (G4EmStandardPhysics_option3) was used

for precise dose calculations and particle-

matter interactions or radiation transport in the

agreement, and

a

reasonable agreement

simulation.

G4EmStandardPhysics_option3

between the measured and simulated dose

distribution were observed with over 99%,

95% and 90% of the points of PDD and cross-

plane profile, respectively.

designed for any applications required higher

accuracy of electrons, hadrons, and ion

tracking without a magnetic field. The package

has been presented in Poon and Arce at al. for

radiotherapy application [27, 28] and

recommended in Varian documents [19]. The

range cuts for gamma, electron, and positron

are fixed to 0.1 mm in a water phantom, 1 mm

in the world volume, and 10 mm in TrueBeam

material volume, respectively.

III. RESULTS AND DISCUSSION

6 PhS files stored 3×10⁸photon,

electron, and positron particles, which have

been simulated. Approximately 6×10⁹particle

histories from 6 PhS files were performed such

that the statistical uncertainty in the dose for

the voxels inside the radiation field was less

than 0.2% at the depth of maximum water

phantom. The simulation results take into

A virtual water phantom with 30×30×30

cm³volume was installed at an SSD equal to

100 cm as the measurement. It is used for the

MC estimation of the absorbed dose

distribution. Voxel size was set to 3×3×3 mm²

for a field size of 10×10 cm².

account

the

electron

and

positron

contamination in the photon beam.

A. PDD curve

E. Gamma Index method

In this paper, the quantity PDD defined

as the quotient, expressed as a percentage, of

the absorbed dose at a predefined depth (dx) to

the absorbed maximum dose at a fixed

reference depth of d0 = 1.5 cm, along the

central axis of the beam. Fig. 2 shows the

comparison between measured and GATE

estimated PDD with SSD = 100 cm for a

10×10 cm²field for FF 6 MV photon beam and

the Gamma Index distribution. The maximum

dose was detected at 1.5 cm of depth in both

measurement and simulation. The statistical

uncertainty of bins scoring PDD was between

0.02% to 0.04% and all bins scoring more than

50% of the maximum absorbed dose was

0.02% to 0.2%. The distribution of Gamma

Data analysis was based on comparisons

between GATE simulations and measurements

using the Gamma Index method [29], which

became a ―gold standard‖ method for the

comparison in dose distribution [13]. This

method was conducted with a percentage dose

difference (ΔD) criteria and distance to

agreement (DTA) of of 1%/1mm and

2%/2mm. If the Gamma Index value is greater

than unity, it indicates a position where the

agreement between the measured and

simulated dose maps do not meet the

predefined criteria. Passing criteria were met if

the gamma index was no larger than 1. An

important feature of this method is that the

40

N. D. TON, B. D. LINH, Q.T. PHAM

Index is shown in Fig. 2, there is only one

dose of the measurement as equation 1. This

discrepancy is never greater than 2%.mm,

2%.

point was larger than 1. The evaluation using

the Gamma Index with 2%/2 mm criteria for

PDD obtained was greater than 98%. There is a

good agreement between the computed and

measured PDD.

|

(1)

Where, D₁and D₂are the value of

simulated and measured, respectively, D₀is the

maximum dose of the measurement.

Fig. 3 shows the percentage dose

difference of the PDD relative to the maximum

Fig.2: Comparison of measured (black line) and GATE simulation estimated (red circle) PDD of TrueBeam

STx FF 6 MV photon beam with SSD =100 cm for 10×10 cm²field. The distribution of gamma index points

of PDD (blue plus).

Fig. 3: Percentage dose difference between the simulated and measured PDD relative to the maximum dose

of the measurement.

41

DOSIMETRIC CHARACTERISTICS OF 6 MV PHOTONS FROM TRUEBEAM STX …

Fig. 4: Comparison of measured (black line) and GATE simulation estimated (red circle) cross-plane profile at 1.5 cm for

TrueBeam STx FF 6 MV photon beam for 10×10 cm²field. The distribution of gamma index points of the cross-plane

profile (blue plus).

Fig. 5: Percentage dose difference between the simulated and measured cross-plane profile at 1.5 cm of depth

relative to the maximum dose of the measurement.

criteria of 1%/1 mm. Although the gamma

indices in the penumbra region (at ±5 cm) are

bigger than those in the inside field, there is

B. Cross-plane Profile

The comparison of profile at 1.5 cm in

the water tank between the simulation and

measure is shown in Fig. 4. The statistical

uncertainty of the simulation is in the range of

0.02% to 0.2%. For the Gamma Index criteria

of 2%/2 mm, the distribution is presented in

Fig. 4 and the average pass rates for the cross-

plane profile was ≥ 94%. This agreement

notably worsens with the more stringent

still

a

reasonable agreement between

computed and measured the cross-plane

profile at 1.5 cm depth.

For cross-plane profiles inside the field

region, the percentage dose difference is less

than 1.5%. The result of the percentage dose

difference is shown in Fig. 5. The differences

42

N. D. TON, B. D. LINH, Q.T. PHAM

for clinical radiotherapy treatment planning,‖

in the penumbra region (at ±5 cm) are bigger

than those in the inside field. These

differences probably represent the number of

simulated particles in which less number of

simulated particles will be found in the

penumbra and result in a big statistical

ﬂuctuations in MC simulation and a big

difference relative [30, 31].

Med. Phys. 25, 1773–1829, 1998.

[2]. N. Reynaert et al., ―Monte Carlo treatment

planning for photon and electron beams,‖

Radiation Physics and Chemistry 76, 643–686,

2007.

[3]. E. Spezi and G. Lewis, ―An overview of

Monte Carlo treatment planning for

radiotherapy,‖ Radiation Protection Dosimetry

131, 123–129, 2008.

IV. CONCLUSIONS

[4]. O. Chibani and C.-M. Ma, ―On the

discrepancies between Monte Carlo dose

calculations and measurements for the 18 MV

Varian photon beam,‖ Med. Phys. 34, 1206–

1216, 2007.

The aim of this work is to validate the

potential application Geant4/GATE software

for the Varian TrueBeam STx. In this study,

the characteristics of 6 MV photons of

TrueBeam STx include PDD and crossline

profile, which was simulated based on

Geant4/GATE using Varian PhS file, and

Varian manufacturer’s information. A PDD

curve and beam profile for 10×10 cm²field

size in a water phantom using Geant4/GATE

simulation show a good agreement with

measured dose data for FF 6 MV photon beam

produced by the Linac. The percentage dose

difference and Gamma Index method were

used for comparison. The agreement between

simulations and experimental data proved that

Geant4/GATE can be used for accurate Monte

Carlo dose estimation.

[5]. M. Constantin et al., ―Modeling the TrueBeam

linac using a CAD to Geant4 geometry

implementation: Dose and IAEA-compliant

Phase Space calculations,‖ Medical Physics

38, 4018–4024, 2011.

[6]. C. Glide-Hurst et al., ―Commissioning of the

Varian TrueBeam linear accelerator: A multi-

institutional study,‖ Medical Physics 40,

031719, 2013.

[7]. G. P. Beyer, ―Commissioning measurements

for photon beam data on three TrueBeam

linear accelerators, and comparison with

Trilogy and Clinac 2100 linear accelerators,‖ J.

Appl. Clin. Med. Phys. 14, 273–288, 2013.

ACKNOWLEDGMENT

[8]. Rodrigues et al., ―A Monte Carlo simulation

framework for electron beam dose calculations

using Varian Phase Space files for TrueBeam

Linacs‖, Medical Physics 42, 2389-2403,

2015.

The authors would like to thank

VINATOM for the support under the grant

number CS/19/04-02, MSc. Nguyen Ngoc

Quynh at INST for his kind help in the run of

Geant4/GATE and all medical physicist at

Department of Radiation Oncology and

Radiosurgery, 108 Military Central Hospital.

[9]. S. Agostinelli et al., ―Geant4—a simulation

toolkit‖, Nucl. Instr. and Meth. in Phys. Res. A

506, 250, 2003.

[10].J. Allison et al., ―Recent developments in

Geant4‖, Nuclear Instruments and Methods in

Physics Research A 835, 186, 2016.

REFERENCES

[1]. B. Fraass et al., ―American Association of

Physicists in Medicine Radiation Therapy

Committee Task Group 53: Quality assurance

[11].L. Grevillot et al., ―Simulation of a 6 MV

Elekta precise linac photon beam using

43

DOSIMETRIC CHARACTERISTICS OF 6 MV PHOTONS FROM TRUEBEAM STX …

GATE/GEANT4, Physics in Medicine &

Biology 56 903–918, 2011.

[22].R. J. Watts et al., ―Accelerator beam data

commissioning equipment and procedures:

Report of the TG-106 of the Therapy Physics

Committee of the AAPM,‖ pp. 4186–4215,

2008.

[12].J. EL Bakkali et al., ―Validation of Monte-

Carlo Geant4 code for Saturne 43 LINAC‖,

Int. J. Innov. Appl. Stud. 4, 424–436, 2013.

[23].S. Jan et al., ―Gate: a simulation toolkit for

PET and SPECT‖, Physics in Medicine &

Biology 49(19), 4543-61, 2004.

[13].S. Didi et al., ―Simulation of the 6 MV Elekta

synergy platform linac photon beam using

Geant4 application for tomographic emission‖,

Medical Physics 40,136–143, 2015.

[24].S. Jan et al, ―GATE V6: a major enhancement

of the GATE simulation platform enabling

modelling of CT and Radiotherapy‖ Physics in

Medicine & Biology 56(4), 881-901, 2011.

[14].Wijesooriya et al., ―Verification of the

TrueBeam Head Shielding Model in Varian

VirtuaLinac via Out-of-field Doses‖, Medical

Physics 46(2), 877-884, 2019.

[25].L. Grevillot et al., ―GATE as a GEANT4-

based Monte Carlo platform for the evaluation

of proton pencil beam scanning treatment

plans‖, Phys. in Med. & Bio. 57(13), 4223-44,

2012.

[15].R. Shende et al., ―Commissioning of

TrueBeam TM Medical Linear Accelerator:

Quantitative and Qualitative Dosimetric

Analysis and Comparison of Flattening Filter

(FF) and Flattening Filter Free (FFF) Beam,‖

no. February, pp. 51–69, 2016.

[26].www.opengatecollaboration.org – accessed:

March 2019.

[16].C. Glide-Hurst et al., ―Commissioning of the

Varian TrueBeam linear accelerator: a multi-

institutional study.,‖ Med. Phys., vol. 40, no. 3,

p. 31719, 2013.

[27].E. Poon and F. Verhaegen, ―Accuracy of the

photon and electron physics in Geant4 for

radiotherapy applications‖, Medical Physics

32, 1696–711, 2005.

[17].G. P. Beyer, ―Commissioning measurements

for photon beam data on three TrueBeam

linear accelerators, and comparison with

Trilogy and Clinac 2100 linear accelerators.,‖

J. Appl. Clin. Med. Phys., vol. 14, no. 1, p.

4077, 2013.

[28].P. Arce, JI lagares, ―CPU time optimization

and precise adjustment of the Geant4 physics

parameters for a VARIAN 2100 C/D gamma

radiotherapy linear accelerator simulation

using GAMOS‖, Physics in Medicine &

Biology 63(3), 035007, 2018.

[18].International

Technical Report IAEA-NDS-0484: ―Phase-

Space Database for External Beam

Atomic

Energy

Agency

[29].D. A. Low et al., ―A technique for the

quantitative evaluation of dose distributions‖,

Medical Physics 25(5), 656-661, 1998.

Radiotherapy‖, IAEA Headquarters, Vienna,

Austria, 2006.

[30].Graves, Y.J., Jia, X. & Jiang, ―Effect of

statistical fluctuation in Monte Carlo based

photon beam dose calculation on gamma index

evaluation‖, Physics in Medicine and Biology

58(6): 1839-1854, 2013.

[19].―TrueBeam Monte Carlo Data Package‖ –

Confidential document for TrueBeam STx

users, version 1.1, 2014.

[20].https://www.iba-dosimetry.com

March 2019.

–

accessed:

[31].Saidi, P., Tenreiro, C. & Sadeghi, ―Variance

Reduction of Monte Carlo Simulation in

Nuclear Engineering Field‖, Rijeka, Croatia:

INTECH Open Access Publisher: Theory and

Applications of Monte Carlo Simulations,

2013.

[21].P. R. Almond and P. J. Biggs, ―AAPM‖ TG-51

protocol for clinical reference dosimetry of

high-energy photon and electron beams,‖ no.

September, pp. 1847–1870, 1999.

44