

ORIGINAL ARTICLE 

Year : 2018  Volume
: 14
 Issue : 3  Page : 647650 

Validation of GATE for bone and bone marrow with calculation specific absorbed fraction for photons
Seydeh Zahra Ghaseminezhad^{1}, Alireza Sadremomtaz^{1}, Hossein Rajabi^{2}
^{1} Department of Physics, University of Guilan, Rasht, Iran ^{2} Department of Medical Physics, School of Medical Sciences, Tarbiat Modares University, Tehran, Iran
Date of Web Publication  12Jun2018 
Correspondence Address: Dr. Alireza Sadremomtaz Department of Physics, University of Guilan, Rasht, Tehran Iran
Source of Support: None, Conflict of Interest: None  Check 
DOI: 10.4103/09731482.183191
Purpose: GATE/GEANT is a Monte Carlo code dedicated to nuclear medicine that allows calculation of the dose to organs (bone and bone marrow) of voxel phantoms. On the other hand, Medical Internal Radiation Dose (MIRD) is a welldeveloped system for estimation of the dose to human organs. In this study, results obtained from GATE/GEANT using leg of Snyder phantom is compared to published MIRD data. Materials and Methods: For this, the mathematical leg of Snyder phantom was discretized and converted to a digital phantom of 100 × 100 × 200 voxels. The activity was considered uniformly distributed within bone and bone marrow. The GATE/GEANT Monte Carlo code was used to calculate the dose to the bone and bone marrow of the leg phantom from monoenergetic photons of 10, 15, 20, 30, 50, 100, 200, 500, and 1000 keV. The dose was converted into a specific absorbed fraction (SAF) and the results were compared to the corresponding published MIRD data. Results: On average, there was a good correlation between the two series of data for selfabsorption (r ^{2} = 0.99) and for crossirradiation (r ^{2} = 0.99). However, the GATE/GEANT data were on average 1.01 ± 0.79% higher than the corresponding MIRD data for selfabsorption. As for crossirradiation, the GATE/GEANT data were on average 8.11 ± 7.95% higher than the MIRD data. Conclusion: In this study, the SAF values derived from GATE/GEANT and the corresponding MIRD published data were compared. On average, the SAF values derived with GATE/GEANT showed an acceptable correlation and agreement with the MIRD data for the photon energies of 50–1000 keV. For photons of 10–30 keV, there was an only poor agreement between the GATE/GEANT results and MIRD data.
Keywords: Bone, bone marrow, GATE, specific absorbed fraction, validation
How to cite this article: Ghaseminezhad SZ, Sadremomtaz A, Rajabi H. Validation of GATE for bone and bone marrow with calculation specific absorbed fraction for photons. J Can Res Ther 2018;14:64750 
How to cite this URL: Ghaseminezhad SZ, Sadremomtaz A, Rajabi H. Validation of GATE for bone and bone marrow with calculation specific absorbed fraction for photons. J Can Res Ther [serial online] 2018 [cited 2020 Jul 8];14:64750. Available from: http://www.cancerjournal.net/text.asp?2018/14/3/647/183191 
> Introduction   
Principally, a newly developed code should be validated against experimental data. However, this is not possible for internal dosimetry, because direct measurement of the dose to the tissues from internally administrated radioisotopes is not possible. In practice, some simple forms of physical phantom are used for validation of the Monte Carlo codes. However, phantom studies are limited to measurement of average dose in a few locations inside the phantoms.^{[1]} Radioisotopes are used in nuclear medicine in a variety of diagnostic and therapeutic procedures. In most applications, a significant absorbed dose may be received by some radiosensitive organs.^{[2]} This requires dose quantification, to balance between the risks and the benefits in any application involving the use of radioisotopes in humans. Unfortunately, direct measurement of the absorbed dose in organs of the human body from administrated radioisotopes is not possible. Therefore, internal dose assessment is often performed using precalculated reference data derived from humanoid anatomical models. For many years, the Medical Internal Radiation Dose (MIRD) Committee of the American society of nuclear medicine has been the main source of reference data for dose assessment in nuclear medicine.^{[3],[4],[5]} A number of Monte Carlo codes are now available for dosimetry applications. Monte Carlo N–Particle Transport Code ^{[6]} and electron gamma shower (EGS)^{[7]} and GEANT4^{[8],[9]} are general purpose codes developed to simulate the interaction of photons and electrons with materials that can be used for internal dose assessment. GATE/GEANT package is the most recently developed Monte Carlo code with the ambition to become the gold standard code in nuclear medicine.^{[10]} GATE/GEANT has certain attractive features for internal dosimetry application.^{[11]} GATE/GENAT is partially validated for imaging applications ^{[12],[13],[14],[15]} and has been used for dosimetry applications ^{[11],[16],[17]} but not validated for nuclear medicine dosimetry in bone and bone marrow. However, MIRD data have never been compared to the data derived from GATE/GEANT for bone and bone marrow. Such comparison is important because GATE/GEANT is the only open access Monte Carlo code dedicated for nuclear medicine that allows calculation of a dose map within the patients' body. On the other hand, MIRD is the most developed system for internal dose assessment in nuclear medicine. In this study, the digital form of the leg of the Snyder mathematical phantom was constructed and used with GATE/GEANT to calculate the dose to bone and bone marrow. The results were compared to the previously published MIRD data.^{[18]}
> Materials and Methods   
Monte Carlo simulation
GATE/GEANT Monte Carlo package (version 4.0.0) was used for estimation of dose to the organs of the phantoms.^{[11]} This version of GATE/GEANT was developed over GEANT4 version 4.9.1.p02. Photoelectric absorption, Compton interaction, Rayleigh scattering, and characteristic Xray production were considered for photon interactions. Xrays were tracked down to 1 keV; below that was assumed absorbed in the same voxel. Each voxel in the phantoms was linked to the table describing the attenuation properties (composition and density) of the corresponding tissues.^{[11]} Simulations were performed for the photons of 10, 15, 20, 30, 50, 100, 200, 500, 1000 keV.
Phantom
To construct a voxel leg phantom identical to the phantom used in the MIRD calculations, the mathematical equations ^{[18]} used to describe bone and bone marrow of the Leg phantom. The Snyder mathematical phantom was sampled at a spatial resolution of 1 mm × 1 mm × 2 mm and converted to a voxel phantom of 100 × 100 × 200 matrix size. The activity of interest was distributed uniformly within bone and bone marrow of the leg phantom, respectively.
Calculation of specific absorbed fraction values
In the MIRD formalism, the reference data are presented in terms of specific absorbed fraction (SAF) values.^{[18],[19],[20]} SAF is a net factor that converts the total energy emitted from a particular source organ to the energy absorbed in another organs (crossirradiation) or the same organ itself (selfabsorption).^{[21]} SAF is defined for each pair of source organ (rs) and target organ (rt) as follows:
where m is the mass of the target organ in kilogram. The absorbed energy in each organ was calculated as sum of the absorbed energy in the entire voxels of the organ. For calculation of organ masses, the total number of voxels belonging to an organ was determined and then multiplied by voxel volume and the density of corresponding organ using MIRD data.^{[18]}
Data analysis
The relative difference (RD) between SAF values derived from GATE/GEANT (SAFGATE) and the corresponding MIRD values (SAFMIRD) for each photon energy was calculated as:
GATE/GEANT and the corresponding MIRD data were compared through fitting a linear curve to the scatter plot of the data and calculating correlation between results.
> Results   
The volumes of the bone and bone marrow in the leg of mathematical Snyder phantom and in its digital from are presented in [Table 1]. RD between the volumes of bone and bone marrow in the two phantoms were 0.03% and 1.5%.  Table 1: Volumes of bone and bone marrow in the mathematical and voxelized leg of Snyder phantom
Click here to view 
SAFs obtained for bone and bone marrow are presented in [Table 2] and [Table 3], respectively. The MIRD SAF values were also included in the tables for the sake of comparison.^{[18]}  Table 2: Specific absorbed fraction values in the leg of Snyder phantom (kg^{1}) for bone as source
Click here to view 
 Table 3: Specific absorbed fraction values in the leg of Snyder phantom (kg^{1}) for bone marrow as source
Click here to view 
Specific absorbed fractions for selfabsorption
[Figure 1] shows the scatter plot and the linear curve fitted to the MIRD and GATE/GEANT data for photons of 10–1000 keV. Due to the wide range of SAF values, the graph is shown in logarithmic scale. The figure clearly shows a good linear correlation (r^{2} = 0.999) between the two series of data. The average and standard deviation of the RDs (%RD) are 1.01 ± 0.79%. We also simulated SAF for the specific energy (140Kev) of the most common radioisotope in nuclear medicine (technesium ^{99m} [^{99m} Tc]). It is 1.2E2 and 3.32E2 for selfabsorption in bone and bone marrow, respectively.  Figure 1: Comparison of GATE and corresponding Medical Internal Radiation Dose specific absorbed fraction values for selfabsorption bone and bone marrow (scatter plot of two series of data)
Click here to view 
Specific absorbed fractions for crossirradiation
The results of a similar analysis for the crossirradiation data are presented in [Figure 2]. The curve fitted to the scatter plot of the data shows a good linear correlation between the SAFGATE/GEANT and the SAFMIRD data. The average and standard deviation of the RDs (%RD) are 8.11 ± 7.95%. For the 140 Kevphoton of ^{99m} Tc, crossirradiation of bone and bone marrow sources is 1.31E2 and 1.1E2, respectively.  Figure 2: Comparison of GATE and corresponding Medical Internal Radiation Dose specific absorbed fraction values for crossirradiated bone and bone marrow (scatter plot of two series of data)
Click here to view 
> Discussion   
Specific absorbed fraction for selfabsorption
The SAF values derived from the voxelized Snyder phantom show a high correlation with the SAFMIRD data. A simple explanation for this observation could be a higher total crosssection and/or density attributed tissue in our study compared to the MIRD calculations. When increasing the mass, less photons can escape from the source organs and therefore, selfabsorption increases.
Specific absorbed fractions for crossirradiation
In some studies, differences between the MIRD SAF values and the corresponding calculated values for lowenergy photons have been reported.^{[4],[22],[23],[24]} However, none of them used the MIRD phantom; therefore, the reported differences may also be due to geometrical differences. Although the RD between our data and the MIRD data are low for highenergy photons (50–1000 keV) and high for lowenergy photons (10–30 keV), this does not imply any direct dependence of the RD on photon energy. When the photon energy is low, most of the photons are absorbed in the source organ, and only a small number of photons penetrate to the target organs. That is, a small fraction of the total energy released is absorbed by the target organs; hence, the corresponding SAF values are small. From the statistical point of view, smaller values show a higher statistical uncertainty (random error). Therefore, the poor agreement between the GATE/GEANT and MIRD data for lowenergy photons is partly due to poor statistics in the corresponding SAF values. In the GATE/GEANT code, required crosssections are derived using the EPDL97, EEDL, and EADL libraries for photons, electrons, and atoms, respectively.^{[25]} The MIRD data were calculated using a Monte Carlo method in which the crosssections were derived from McMaster tables and the ENDL library.^{[11],[26],[27]} For photons of 50–1000 keV, different types of interactions in multiple stages are possible. The absorbed energy, therefore, depends on a random combination of crosssections that may cancel out pointbypoint differences. However, for lowenergy photons (10–30 keV), where the photoelectric interaction is predominant, and the average number of interactions along the particle track is small, the difference in the crosssections becomes evident. This may be one reason for the larger difference between the present GATE/GEANT and the MIRD data, for lowenergy photons.
> Conclusion   
In this study, the SAFGATE/GEANT and the SAFMIRD published data were compared. On average, the SAF values derived by GATE/GEANT showed an acceptable correlation and agreement with the MIRD data for the photon energies of 50–1000 keV. For photons of 10–30 keV, there was an only poor agreement between the GATE/GEANT results and MIRD data. To some extent, this low agreement was due to differences in crosssection tables used in GATE/GEANT and MIRD. In general, the agreement was acceptable, and the results can be considered as validation of GEANT/GATE against MIRD. It is worth to mention that although these calculations are easy to use, it can differ from the human body because it is based on a specific model of the human phantom.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
> References   
1.  Kron T, Duggan L, Smith T, Rosenfeld A, Butson M, Kaplan G, et al. Dose response of various radiation detectors to synchrotron radiation. Phys Med Biol 1998;43:323559. [ PUBMED] 
2.  Stabin MG. Developments in the internal dosimetry of radiopharmaceuticals. Radiat Prot Dosimetry 2003;105:57580. [ PUBMED] 
3.  Chao TC, Xu XG. Svalues calculated from a tomographic head/brain model for brain imaging. Phys Med Biol 2004;49:497184. [ PUBMED] 
4.  Ferrari P, Gualdrini G. MCNPX internal dosimetry studies based on the NORMAN05 voxel model. Radiat Prot Dosimetry 2007;127:20913. [ PUBMED] 
5.  Wessels BW, Syh JH, Meredith RF. Overview of dosimetry for Systemic targeted radionuclide therapy (STaRT). Int J Radiat Oncol Biol Phys 2006;66 2 Suppl: S3945. 
6.  Briesmeister JF. MCNP – A General Monte Carlo N – Particle Transport Code. Version 4C. Los Alamos National Laboratory Report 2000. 
7.  Kawrakow I, Rogers DW. The EGSnrc Code System: Monte Carlo Simulation of Electron and Photon Transport: Ionizing Radiation Standards National Research Council of Canada. Report No. NRCC PIRS701; 2000. 
8.  Guimarães CC, Moralles M, Okuno E. Performance of GEANT4 in dosimetry applications: Calculation of xray spectra and kermatodose equivalent conversion coefficients. Radiat Meas 2008;43:152531. 
9.  Rodrigues P, Trindade A, Peralta L, Alves C, Chaves A, Lopes MC. Application of GEANT4 radiation transport toolkit to dose calculations in anthropomorphic phantoms. Appl Radiat Isot 2004;61:145161. [ PUBMED] 
10.  Santin G, Staelens S, Taschereau R, Descourt P, Schmidtlein CR, Simon L, et al. Evolution of the GATE project: New results and developments. Nucl Phys B (Proc Suppl) 2007;172:1013. 
11.  Visvikis D, Bardies M, Chiavassa S, Danford C, Kirov A, Lamare F, et al. Use of the GATE Monte Carlo package for dosimetry applications. Nucl Instrum Methods A 2006;569:33540. 
12.  Assié K, Gardin I, Véra P, Buvat I. Validation of the Monte Carlo simulator GATE for indium111 imaging. Phys Med Biol 2005;50:311325. 
13.  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;33:198208. [ PUBMED] 
14.  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;49:454361. 
15.  Staelens S, Strul D, Santin G, Vandenberghe S, Koole M, D'Asseler Y, et al. Monte Carlo simulations of a scintillation camera using GATE: Validation and application modelling. Phys Med Biol 2003;48:302142. 
16.  Ferrer L, Chouin N, Bitar A, Lisbona A, Bardiès M. Implementing dosimetry in GATE: Dosepoint kernel validation with GEANT4 4.8.1. Cancer Biother Radiopharm 2007;22:1259. 
17.  Taschereau R, Chatziioannou AF. FDGPET imagebased dose distribution in a realistic mouse phantom from Monte Carlo simulations. IEEE Nucl Sci Symp Conf Rec 2005;3:163336. 
18.  Snyder WS, Ford MR, Warner GG. Estimates of absorbed fractions for monoenergetic photon sources uniformly distributed in various organs of a heterogeneous phantom. MIRD Pamphlet No. 5. New York, USA: Society of Nuclear Medicine; 1978. 
19.  Blaickner M, Kindl P. Diversification of existing reference phantoms in nuclear medicine: Calculation of specific absorbed fractions for 21 mathematical phantoms and validation through dose estimates resulting from the administration of (18)FFDG. Cancer Biother Radiopharm 2008;23:76782. [ PUBMED] 
20.  Smith T, PetoussiHenss N, Zankl M. Comparison of internal radiation doses estimated by MIRD and voxel techniques for a “family” of phantoms. Eur J Nucl Med 2000;27:138798. [ PUBMED] 
21.  Sgouros G. Dosimetry of internal emitters. J Nucl Med 2005;46 Suppl 1:18S27S. [ PUBMED] 
22.  Larsson E, Jönsson BA, Jönsson L, Ljungberg M, Strand SE. Dosimetry calculations on a tissue level by using the MCNP4c2 Monte Carlo code. Cancer Biother Radiopharm 2005;20:8591. 
23.  Kolbert KS, Sgouros G, Scott AM, Bronstein JE, Malane RA, Zhang J, et al. Implementation and evaluation of patientspecific threedimensional internal dosimetry. J Nucl Med 1997;38:3018. [ PUBMED] 
24.  Zankl M, PetoussiHenss N, Fill U, Regulla D. The application of voxel phantoms to the internal dosimetry of radionuclides. Radiat Prot Dosimetry 2003;105:53948. [ PUBMED] 
25.  Jan S, Santin G, Strul D, Staelens S, Assi´e K, Autret D, et al. GATE: A simulation toolkit for PET and SPECT. Phys Med Biol 2004;49:454361. 
26.  Mcmaster WH, Del Grande NK, Mallett JH, Hubbell JH. Compilation of XRay Cross Sections, Ucrl50174 Sec. ii, Rev i. Livermore: University of California, Lawrence Radiation Laboratory; 1969. p. 261. 
27.  Plechaty EF, Terrall JR. An integrated system for production of neutronics and photonics calculational constants. Vol. VI. Photon cross sections 1 Kev to 100 Mev. UNCL. Orig. Receipt Date: 31, December 69, Other Information; 1968. 
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3]
