- Journal List
- J Med Signals Sens
- v.4(4); Oct-Dec 2014
- PMC4236805
As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsem*nt of, or agreement with, the contents by NLM or the National Institutes of Health.
Learn more: PMC Disclaimer | PMC Copyright Notice
J Med Signals Sens. 2014 Oct-Dec; 4(4): 267–273.
PMCID: PMC4236805
PMID: 25426430
Mohammad-Taghi Bahreyni-Toosi, Shahrokh Nasseri, Mahdi Momennezhad, Fatemeh Hasanabadi, and Hamid Gholamhosseinian
Author information Article notes Copyright and License information PMC Disclaimer
Abstract
The aim of this study is to provide a control software system, based on Monte Carlo simulation, and calculations of dosimetric parameters of standard and wedge radiation fields, using a Monte Carlo method. GATE version 6.1 (OpenGATE Collaboration), was used to simulate a compact 6 MV linear accelerator system. In order to accelerate the calculations, the phase-space technique and cluster computing (Condor version 7.2.4, Condor Team, University of Wisconsin–Madison) were used. Dosimetric parameters used in treatment planning systems for the standard and wedge radiation fields (10 cm × 10 cm to 30 cm × 30 cm and a 60° wedge), including the percentage depth dose and dose profiles, were measured by both computational and experimental methods. Gamma index was applied to compare calculated and measured results with 3%/3 mm criteria. Gamma index was applied to compare calculated and measured results. Almost all calculated data points have satisfied gamma index criteria of 3% to 3 mm. Based on the good agreement between calculated and measured results obtained for various radiation fields in this study, GATE may be used as a useful tool for quality control or pretreatment verification procedures in radiotherapy.
Keywords: Clustering, GEANT4 Application for Tomographic Emission, Monte Carlo simulation, phase-space technique, wedge radiation fields
INTRODUCTION
Access to accurate treatment planning is required to obtain a reliable dose distribution. Using simulation models as a standard tool in the optimization of software systems, seems to be efficient and economical. Monte Carlo method is one of the most accurate dosimetry techniques; among the available codes in this method, GATE has a high degree of acceptance among researchers.[1,2,3]
GATE code, presented in 2004, is a subset of GEANT4 Monte Carlo code.[4] In the first place, this code was specifically designed for the simulation of nuclear medicine devices;[1,5,6,7,8,9,10,11,12,13] though recently it has also been used for radiation Therapy tasks[14,15,16,17,18,19,20] and computed tomography due to its flexibility.[21,22,23]
The goal of this project is to provide a software-based control system, for the optimization of dosimetric parameters of LINAC systems, used in radiotherapy centers. To achieve this goal, a 6 MV photon beam of compact linear accelerator (Elekta, Stockholm, Sweden) was simulated using the GATE code. In addition, the analysis of dosimetric parameters of photon beams, such as percentage depth dose and dose profiles in the water phantom for the standard and wedge radiation fields was performed. The findings were compared with the corresponding experimental results.
MATERIALS AND METHODS
Software and hardware requirements for this study were as follows:
Hardware Requirements
The 6 MV photon beam used in this study, was delivered by Elekta linear accelerator. Dosimetry was performed according to TG-51 protocol.[24] To collect data, a water phantom as well as a diode detector (Wellhoffer – Scanditronix, Schwarzenbruck, Germany) was applied.
In order to accelerate the simulation calculations, the parallel computing technique was used on 9 computers (Intel (R) core (TM) 2 Duo CPU with 2.93 GHz, 2GB RAM).
Software Requirements
On the way to perform the simulation, GEANT4 and GATE codes were used (versions 4.9.3 and 6.1, respectively). In order to store/analyze the data during particle simulation, ROOT version 5.27.4, which is an object-oriented data analysis framework, was used.[25] To view and verify the geometry implementation, graphical interfaces, which are available in GEANT4/GATE, were applied. These interfaces were as follows: WIRED 3,[26] VRML viewer 4.0,[27] and DAWN version 3_88.[28] The operating systems used in the study were Fedorc core 13 and CentOs,[29] version 6.0.[30] For parallel computing and clustering, Condor (platform) version 7.2.4 was used.[31]
Geometrical Implementation of Compact Linear Accelerator System
Physical characteristics (shape, geometric dimensions and the material of constituent elements) of the original compact linear accelerator treatment head was defined in the GATE Code, which included the target: Made of tungsten alloy, about 0.2 cm thickness, the primary collimator: Made of tungsten, 10.2 cm height, located below the X-ray target used to collimate the X-ray in the direction of the treatment field, the flattening filter: Made of stainless steel and conical shape and its height is 17.5 mm and 2 mm in the middle and corner, respectively, the ionization chamber, a 60-degree universal wedge, and the secondary collimators: Are made of tungsten alloy about 10 cm thickness. Figure 1 is the view of the linear accelerator system simulation.
Figure 1
The view of the linear accelerator system
The Definition of Electron Source
For defining the electron beam incident on the target of linear accelerator system, the general particle source module was used.[32] This module provides an opportunity to define and implement parameters such as spatial and angular distributions and the energy spectrum of the electron beam. Trial and error method was used to determine these parameters. After setting these parameters, the dose distributions, calculated in the water phantom, were compared with the experimental data using the gamma index method[33] with 3%/3 mm criteria.
The Definition of Physical Interactions
The exact implementation of a system like LINAC requires the simulation of all physical events, which occur in the real world. Considering the energy range of the primary and secondary particles, which are produced in a linear accelerator, the standard model of electromagnetic interactions was utilized. Seven major categories of physics processes are provided by GEANT4. The following is a list of the standard electromagnetic processes available in Geant4: Photon processes, Electron/positron processes, Muon processes, Hadron/ion processes, Coulomb scattering processes, Processes for simulation of polarized electron and gamma beams, Processes for simulation of X-rays and optical protons production by charged particles.[34]
Application of the Actors
Actors are tools that let to interact with GATE. With the aim of extracting the dosimetric parameters in radiotherapy, the Actors should be used in GATE simulation process.[35] DoseActor and KillActor are used in the calculation of dosimetric parameters and acceleration of the simulation process, respectively.
Implementation Stages of Simulation
Stage 1: Defining the phase-space, tracking the primary and secondary particles, and recording information about the particles passing through the phase space.
At this stage of the simulation, the primary particles were electrons. All the primary and secondary particles passing through the phase space, under the flattening filter, were recorded.
KillActor was employed to accelerate the simulation process. As it can be seen in Figure 2, the particles tracking are confined to regions where they are actually influential on the dosimetric parameters in the water phantom.
Figure 2
The view of the particles trajectory, using KillActor
Stage 2: Tracking the exit photons of the phase space, calculating dose distributions, and recording the dosimetric parameters.
At this stage, the primary particles are the same particles produced in the first phase-space stage. The components of LINAC that are present in the trajectory of particles, from the phase space to the phantom water, include the wedge (in wedge fields) and secondary collimator.
The particles trajectory from the phase space to the water phantom is shown in Figure 3.
Figure 3
The trajectory of particles from the phase space and the incident on the opaque water phantom
Clustering
With the purpose of accelerating the calculations, the cluster computing technique (Condor, platform, version 7.2.4) was utilized, and Condor was used on 9 computers (Intel (R) core (TM), 2 Duo CPU with 2.93 GHz, 2GB RAM).
RESULTS
The results of this study include the computational and experimental dosimetric parameters. To ensure the accuracy of the simulation results, it is necessary to analyze the correctness of the simulation process. Therefore, prior to calculating the dose distribution in the water phantom, the energy spectrum, the spatial distribution of electron beams, and the implementation of the linear accelerator system were verified.
Evaluation of Electron Beam Characteristics
Initially, the trial and error method was used to characterize the electron beam incident on the target surface, and the electron energy spectrum was used in the calculation of dosimetry parameters in the water phantom. The specifications of the electron beam energy and the angular direction, used in the calculation of two- and three-dimensional dose distributions in the water phantom, included two half Gaussian curves with the mean energy of 6.2 MeV, and the standard deviations of 0.2 and 0.3 MeV above/below average, respectively, and one-dimensional accelerator beam (beam1d) with the standard deviation of 1.65;[32] as shown in Figure 4. Two- and three-dimensional spatial distribution of incident electrons on the target surface are demonstrated in Figures Figures44 and and5.5. These figures were plotted by ROOT framework.
Figure 4
The electron energy spectrum in the target surface
Figure 5
Three-dimensional distributions of electron beam in the target surface
Overall Evaluation of the Geometry of the Linear Accelerator
To evaluate and verify the implementation of the radiation field and the geometry of linear accelerator systems, particularly the secondary collimator, the radiation flux of particles at SSD = 100 cm, and the data about the particles in the phase space were recorded. The results of the implementation of LINAC system for the radiation field of 10 × 10 cm2 is presented in Figure 6. This figure was plotted by ROOT framework.
Figure 6
Three-dimensional and two-dimensional distributions of the radiation flux of particles at SSD = 100 cm, for 10 × 10 cm2 (a) and (b) radiation field
The geometric accuracy of the implementation of the secondary collimator system is apparent in the resultant graphs. Some properties of the particles passing through the phase space below the flattening filter were evaluated using the saved ROOT file; these characteristics are such as the type and energy of the particle, and the unit vector components corresponding to the particle movement direction. Afterwards, the energy spectrum, the spatial distribution of the coordinates (X, Y, Z), and the unit vectors (signifying directions) (dX, dY, dZ) were drawn. The distributions of (dX, dY) and (X, Y) were similar, and the results confirmed the accuracy of simulation program.
Results of Dosimetric Parameters and the Gamma Index
After ensuring the accuracy of the simulation program, the computational and experimental results of the dosimetric parameters, such as the percentage depth dose and profile dose, for standard and wedge radiation fields (size = 10 × 10 to 30 × 30 cm), were drawn as curves; gamma index was used to compare the computational and empirical results.
In the first stage of the simulation, parallel computing technique, on 26 CPU nodes, was employed to create ROOT files that contain specifications of the particles that are passing through the phase space. On each node, 125,000,000 electrons were tracked, and 26 ROOT files were stored, with 1.2-1.5 gigabyte capacity. In the second stage, the root files were used to calculate the three-dimensional absorbed dose distribution in the water phantom. At this stage, 200,000,000 particles were tracked on each node; notably, the calculation error was <1%. Furthermore, results have satisfied gamma index criteria 3%/3 mm. For instance, the experimental and calculated results of the percentage depth dose, along with the gamma index, for the open radiation fields (10 × 10 and 30 × 30 cm), are demonstrated in Figure 7.
Figure 7
The curves of the percentage depth dose and Gamma index for the open radiation fields 10 cm × 10 cm (a) and 30 cm × 30 cm (b), Blue dots: Calculated data, red dots: Measured data
Also, the results of computational and experimental dose profiles, along with the gamma index, for the radiation fields are shown in Figure 8. In this figure there are some points that gamma index is higher than 1, These points are out of radiation fields, So doses are very low and a little change of calculated dose leads to high percentage difference between calculated and measured dose.
Figure 8
The curves of the dose profile and Gamma Index for the open radiation fields 10 cm ×10 cm (a) and 30 cm × 30 cm (b)
The curves of the percentage depth dose, and profile dose (along with gamma index), for the 60º wedge radiation filed (10 × 10 cm), are presented in Figures Figures9a9a and andb,b, respectively.
Figure 9
The curves of the percentage depth dose (a), Dose profile (b) and related Gamma Index for the 60° wedge radiation field 10 cm ×10 cm
Three-dimensional Dose Distribution Images
As noted in previous sections, DoseActor was used to calculate the absorbed dose, deposited energy, computational errors, and the number of hits in the water phantom. This actor was attributed to the total volume of the water phantom. The outputs of the DoseActor are images with analyze format and two files with.hdr and.img extensions. By using DoseActor, transverse images of a 3-dimensional matrix of the aforementioned parameters, with a voxel size of 5 × 5 × 5, can be presented.
The coronal images of the open and 60º wedge radiation fields (10 × 10 cm) are shown in Figure 10a and andb,b, respectively.
Figure 10
The Coronal sections of the three-dimensional absorbed dose in the water phantom, irradiated by a 10 cm × 10 cm open (a) and 60° wedge (b) Radiation field
DISCUSSION
The purpose of this study is to simulate the compact linear accelerator system and to provide a software-based dosimetry system, according to Monte Carlo calculations and GATE computational code. In this study, the simulation of the geometric components of the system was designed with a precision of 0.01 mm. The geometry of the simulated linear accelerator system was evaluated by the graphical drivers, included in GEANT4/GATE.
Since full tracking of all the particles (primary and secondary), and recording of the dosimetric parameters (such as the three-dimensional absorbed dose distribution) in a certain space of the world volume is time-consuming, the phase space was used for accelerating the simulation. Also, for improving the simulation, the technique of parallel computing was utilized.
As can be seen in Figures Figures8a8a and andb,b, there is only a minor statistical difference between the experimental measurements and the data obtained with the GATE simulations for profile dose curves (up to 1.9% and 1.6% for 10 × 10 cm and 30 × 30 cm, respectively).
There is bigger difference between measured and calculated dose in 30 × 30 cm compared to 10 × 10 cm, because Field size 10 × 10 cm is reference field and dosimetric properties in simulation primarily set in this field. It should be noted both fields have acceptable gamma index in a flat region of treatment fields. Based on the good agreement between calculated and measured results obtained for various radiation fields in this study, GATE may be used as a useful tool for evaluation of quality control in radiotherapy.
Today with the advances in treatment planning system for conformal therapy, it is essential to have isodose curves of the open and wedge radiation fields. Using these curves in the radiotherapy departments prevent the interruption of treatments. Besides, the ability of GATE code to calculate 3-D absorbed dose distribution can help with the calculation of these curves.
BIOGRAPHIES
Mohammad-Taghi Bahreyni-Toosi Professor in medical physics, Medical Physics Research Center, Medical Physics Department, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran.
E-mail:ri.ca.smum@TMinyerhab
Shahrokh Nasseri Assistant professor in medical physics, Medical Physics Research Center, Medical Physics Department, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran.
E-mail:ri.ca.smum@HSiresan
Mahdi Momennezhad Professor assistant in medical physics, Medical Physics Research Center, Medical Physics Department, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran.
E-mail:ri.ca.smum@Mdahzennemom
Fatemeh Hasanabadi M.Sc degree in medical physics from Mashhad University of Medical Sciences, Mashhad, Iran (2014).
E-mail:ri.ca.smum@198Fidabanasah
Hamid Gholamhosseinian received B.Sc degree in radiation therapy technology from Tehran University of Medical Science, Tehran, Iran (1996), M.Sc degree in medical physics (2002) and Ph.D. degree in medical physics from Mashhad University of Medical Sciences, Mashhad, Iran(2014). His research interest is Monte Carlo simulation in treatment planning, dosimetry and radiation therapy physics.
E-mail:ri.ca.smum@1hnainiessohmalohg
ACKNOWLEDGMENTS
We would like to thank Elekta Medical Systems, for providing detailed information on Elekta Compact linear accelerators.
Footnotes
Source of Support: This work is extracted from a PhD thesis that was supported by the vice chancellor for research of Mashhad University of Medical Sciences grant No. 911033
Conflict of Interest: None declared
REFERENCES
1. Jan S, Santin G, Strul D, Staelens S, Assié K, Autret D, et al. GATE: A simulation toolkit for PET and SPECT. Phys Med Biol. 2004;49:4543–61. [PMC free article] [PubMed] [Google Scholar]
2. Jabbari K, Anvar HS, Tavakoli MB, Amouheidari A. Monte Carlo Simulation of Siemens ONCOR Linear Accelerator with BEAMnrc and DOSXYZnrc Code. J Med Signals Sens. 2013;3:172–9. [PMC free article] [PubMed] [Google Scholar]
3. Jabbari K, Roayaei M, Saberi H. Calculation of excess dose to the eye phantom due to a distanced shielding for electron therapy in head and neck cancers. J Med Signals Sens. 2012;2:144–8. [PMC free article] [PubMed] [Google Scholar]
4. Allison J, Amako K, Apostolakis J, Araujo H, Dubois PA, Asai M, et al. Geant4 developments and applications. IEEE Trans Nucl Sci. 2006;53:270–8. [Google Scholar]
5. Assié K, Gardin I, Véra P, Buvat I. Validation of the Monte Carlo simulator GATE for indium-111 imaging. Phys Med Biol. 2005;50:3113–25. [PubMed] [Google Scholar]
6. Bruyndonckx P, Lemaître C, Schaart D, Maas M, van der Laan D, Krieguer M, et al. Towards a continuous crystal APD-based PET detector design. Nucl Instrum Methods Phys Res A. 2007;571:182–6. [Google Scholar]
7. Chung YH, Choi Y, Cho G, Choe YS, Lee KH, Kim BT. Optimization of dual layer phoswich detector consisting of LSO and LuYAP for small animal PET. IEEE Trans Nucl Sci. 2005;52:217–21. [Google Scholar]
8. Gonias P, Bertsekas N, Karakatsanis N, Saatsakis G, Gaitanis A, Nikolopoulos D, et al. Validation of a GATE model for the simulation of the Siemens biograph™ 6 PET scanner. Nucl Instrum Methods Phys Res A. 2007;571:263–6. [Google Scholar]
9. Karakatsanis N, Sakellios N, Tsantilas N, Dikaios N, Tsoumpas C, Lazaro D, et al. Comparative evaluation of two commercial PET scanners, ECAT EXACT HR+and Biograph 2, using GATE. Nucl Instrum Methods Phys Res A. 2006;569:368–72. [Google Scholar]
10. Lamare F, Turzo A, Bizais Y, Le Rest CC, Visvikis D. Validation of a Monte Carlo simulation of the Philips Allegro/GEMINI PET systems using GATE. Phys Med Biol. 2006;51:943–62. [PubMed] [Google Scholar]
11. Schmidtlein CR, Kirov AS, Nehmeh SA, Erdi YE, Humm JL, Amols HI, et al. Validation of GATE Monte Carlo simulations of the GE Advance/Discovery LS PET scanners. Med Phys. 2006;33:198–208. [PubMed] [Google Scholar]
12. van der Laan D, Maas M, De Jong H, Schaart D, Bruyndonckx P, Lemaître C, et al. Simulated performance of a small-animal PET scanner based on monolithic scintillation detectors. Nucl Instrum Methods Phys Res A. 2007;571:227–30. [Google Scholar]
13. Visvikis D, Lefevre T, Lamare F, Kontaxakis G, Santos A, Darambara D. Monte Carlo based performance assessment of different animal PET architectures using pixellated CZT detectors. Nucl Instrum Methods Phys Res A. 2006;569:225–9. [Google Scholar]
14. Benhalouche S, Visvikis D, Le Maitre A, Pradier O, Boussion N. Evaluation of clinical IMRT treatment planning using the GATE Monte Carlo simulation platform for absolute and relative dose calculations. Med Phys. 2013;40:021711. [PubMed] [Google Scholar]
15. Grevillot L, Bertrand D, Dessy F, Freud N, Sarrut D. A Monte Carlo pencil beam scanning model for proton treatment plan simulation using GATE/GEANT4. Phys Med Biol. 2011;56:5203–19. [PubMed] [Google Scholar]
16. Grevillot L, Bertrand D, Dessy F, Freud N, Sarrut D. GATE as a GEANT4-based Monte Carlo platform for the evaluation of proton pencil beam scanning treatment plans. Phys Med Biol. 2012;57:4223–44. [PubMed] [Google Scholar]
17. Grevillot L, Frisson T, Maneval D, Zahra N, Badel JN, Sarrut D. Simulation of a 6 MV Elekta Precise Linac photon beam using GATE/GEANT4. Phys Med Biol. 2011;56:903–18. [PubMed] [Google Scholar]
18. Maigne L, Perrot Y, Schaart DR, Donnarieix D, Breton V. Comparison of GATE/GEANT4 with EGSnrc and MCNP for electron dose calculations at energies between 15 keV and 20 MeV. Phys Med Biol. 2011;56:811–27. [PubMed] [Google Scholar]
19. Thiam CO, Breton V, Donnarieix D, Habib B, Maigne L. Validation of a dose deposited by low-energy photons using GATE/GEANT4. Phys Med Biol. 2008;53:3039–55. [PubMed] [Google Scholar]
20. Sadoughi HR, Nasseri S, Momennezhad M, Sadeghi HR, Bahreyni-Toosi MH. A Comparison Between GATE and MCNPX Monte Carlo Codes in Simulation of Medical Linear Accelerator. J Med Signals Sens. 2014;4:10–7. [PMC free article] [PubMed] [Google Scholar]
21. Brunner FC, Khoury R, Benoit D, Meessen C, Bonissent A, Morel C. Simulation of PIXSCAN, a photon counting micro-CT for small animal imaging. J Instrum. 2009;4:P05012. [Google Scholar]
22. Chen Y, Liu B, O’Connor JM, Didier CS, Glick SJ. Characterization of scatter in cone-beam CT breast imaging: Comparison of experimental measurements and Monte Carlo simulation. Med Phys. 2009;36:857–69. [PMC free article] [PubMed] [Google Scholar]
23. Nicol S, Karkar S, Hemmer C, Dawiec A, Benoit D, Breugnon P, et al., editors. Design and Construction of the ClearPET/XPAD Small Animal PET/CT Scanner. Nuclear Science Symposium Conference Record (NSS/MIC), 2009 IEEE. 2009 IEEE. [Google Scholar]
24. Almond PR, Biggs PJ, Coursey BM, Hanson WF, Huq MS, Nath R, et al. AAPM's TG-51 protocol for clinical reference dosimetry of high-energy photon and electron beams. Med Phys. 1999;26:1847–70. [PubMed] [Google Scholar]
25. Antcheva I, Ballintijn M, Bellenot B, Biskup M, Brun R, Buncic N, et al. ROOT – A C++ framework for petabyte data storage, statistical analysis and visualization. Comput Phys Commun. 2009;180:2499–512. [Google Scholar]
26. Geant4 Vis Tutorial using the WIRED Event Display. [Last cited on 2013 Nov 04]. Available from:http://www.conferences.fnal.gov/g4tutorial/g4cd/Documentation/Visualization/G4WIREDTutorial/G4WIREDTutorial.html .
27. Allison J, Asai M, Barrand G, Donszelmann M, Minamimoto K, Perl J, et al. The geant4 visualisation system. Comput Phys Commun. 2008;178:331–65. [Google Scholar]
28. Tanaka S, Kawaguti M, editors. DAWN for Geant4 Visualization. Proceedings of the CHEP. 1997 [Google Scholar]
29. CentOS: The Community ENTerprise Operating System. [Last cited on 2013 Nov 04]. Available from:http://www.centos.org/
30. Fedora Project. [Last cited on 2013 Nov 04]. Available from:http://www.fedoraproject.org/
31. Computing with HTCondor. [Last cited on 2013 Nov 04]. Available from:http://www.research.cs.wisc.edu/htcondor/
32. Ferguson C. General Purpose Source Particle Module for GEANT4/SPARSET: Technical Note. Uos-GSPM-Tech. 2000. [Last cited on 2013 Nov 01]. Available from:http://www.reat.space.qinetiq.com/gps/gspm_docs/gspm_tn1 pdf .
33. Chen M, Lu W, Chen Q, Ruchala K, Olivera G. Efficient gamma index calculation using fast Euclidean distance transform. Phys Med Biol. 2009;54:2037–47. [PubMed] [Google Scholar]
34. Users Guide: Setting up the physics. [Last cited on 2013 Nov 04]. Available from:http://www.wiki.opengatecollaboration.org/index.php/Users_Guide: Setting_up_the_physics .
35. Users Guide V6:Readout parameters for Radiotherapy applications: Actors. [Last cited on 2013 Sep 21]. Available from:http://www.wiki.opengatecollaboration.org/index.php/Users_Guide_V6:Readout_parameters_for_Radiotherapy_applications: _Actors .
Articles from Journal of Medical Signals and Sensors are provided here courtesy of Wolters Kluwer -- Medknow Publications
