- Email: [email protected]
Performance evaluation of the source description of the THOR BNCT epithermal neutron beam
Applied Radiation and Isotopes 69 (2011) 1892–1896
Contents lists available at ScienceDirect
Applied Radiation and Isotopes journal homepage: www.elsevier.com/locate/apradiso
Performance evaluation of the source description of the THOR BNCT epithermal neutron beam Yuan-Hao Liu a,n, Pi-En Tsai a, Hui-Ting Yu b, Yi-Chun Lin c, Yu-Shiang Huang b, Chun-Kai Huang b, Yen-Wan Hsueh Liu b, Hong-Ming Liu a, Shiang-Huei Jiang b a
Nuclear Science and Technology Development Center, National Tsing Hua University, Hsinchu, Taiwan Institute of Nuclear Engineering and Science, National Tsing Hua University, HsinChu, Taiwan c Department of Biomedical Engineering and Environmental Sciences, National Tsing Hua University, HsinChu, Taiwan b
a r t i c l e i n f o
a b s t r a c t
Available online 13 April 2011
This paper aims to evaluate the performance of the source description of the THOR BNCT beam via different measurement techniques in different phantoms. The measurement included (1) the absolute reaction rate measurement of a set of triple activation foils, (2) the neutron and gamma-ray dose rates measured using the paired ionization chamber method, and (3) the relative reaction rate distributions obtained using the indirect neutron radiography. Three source descriptions, THOR-Y09, surface source file RSSA, and THOR-50C, were tested. The comparison results concluded that THOR-Y09 is a welltested source description not only for neutron components, but also for gamma-ray component. & 2011 Elsevier Ltd. All rights reserved.
Keywords: THOR BNCT Epithermal neutron beam Source description
1. Introduction In 2004, the Tsing Hua Open-pool Reactor (THOR) in Hsinchu, a 2 MW TRIGA reactor, was modified into a boron neutron capture therapy (BNCT) facility with an intense epithermal neutron beam, reaching 1.07 109 epithermal neutrons cm 2 s 1 at 1.2 MW (Liu et al., 2009a). For BNCT, accurately predicting dose delivery depends not only on determination of boron distribution and concentration but also on validness of the treatment planning system. Hence, a source description plays a very important role in treatment planning system; it is necessary to validate the implemented source description of the used BNCT beam to ensure the correctness of calculated dose. Correspondingly, it is of our interest to evaluate the performance of the THOR source description in different scenarios. To achieve such a purpose, a series of delicate beam dosimetry, characterization, and adjustment have been performed; activation detectors, indirect neutron radiography, and paired ionization chambers were employed as measuring tools. Measurements were conducted in different PMMA phantoms. The following sections will detail the evaluation work.
2. Materials and methods 2.1. Source description Generally, a source description consists of neutron and gamma-ray components. At THOR, the generalized source n
Corresponding author. Tel.: þ886 3 5715131x42889. E-mail addresses: [email protected], [email protected] (Y.-H. Liu).
0969-8043/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2011.04.005
description, denoted as THOR-Y09, includes four sets of sources: fast neutron, epithermal neutron, thermal neutron, and gamma-ray sub-sources. Each sub-source has its own energy, angular, and spatial distributions. THOR-Y09 was an experimentally adjusted version of the calculated, generalized source description THOR-50C, which has only two sets of sources: one neutron and one gamma-ray sub-sources. The neutron source adjustment utilized full-scale deconvolution; this innovative methodology not only deconvoluted the energy distribution but also adjusted the angular and spatial distributions (Liu, 2009b; Liu et al., 2010). The gamma-ray source was adjusted through a multi-cap magnesium-argon ionization chamber with the aid of Monte Carlo calculated detector response functions (Liu et al., this issue-a). The source surfaces of THOR-50C and THOR-Y09 are both located at the irradiation port aperture surface. In addition to THOR-Y09 and THOR-50C, a surface source RSSA was created using the Monte Carlo code MCNP5 (X-5 Monte Carlo Team, 2005) with a relatively modified/improved model of reactor core and beam configuration compared to the one used in generating THOR-50C. Surface source distribution has a significant advantage that its azimuthal angle information is more faithful than generalized source description (Albritton and Kiger, 2008). However, due to the huge amount of data it records to achieve trustiness, the RSSA is not handy to use. For example, in our test, RSSA file cannot be used via MPI interface to make parallel calculation; the applied RSSA file has a size of 60 GB. Furthermore, when a discrepancy is found between measurement and calculation, it is important that the source is adjustable, which is unavailable for the RSSA file. Unfortunately, this is the case at THOR, despite the fact that with proper variance reduction
Y.-H. Liu et al. / Applied Radiation and Isotopes 69 (2011) 1892–1896
1893
techniques the size of RSSA could be effectively reduced (Albritton and Kiger, 2010). Nevertheless, using a customized version of MCNP5 such as the one modified by Albritton and Kiger (2008), the RSSA file can be used as a set of default parameters built outside MCNP. Until present, however, the modified version is not yet available at THOR but will be considered in the future work. No matter how delicate a source description is, it has to be proven reliable and trustworthy. To prove the trustiness of THOR-Y09, tests were carried out in this work. 2.2. Measurement tools The triple-foil set plays a role as activation detectors to measure the neutron intensity in the study (Chang et al., 2006; Tsai et al., 2010). It consists of AuAl (1 wt% Au, 0.2 mm thick), natural Cu (0.1 mm thick), and MnNi (88 wt% Mn, 0.1 mm thick) foils in sequence. Foil diameter is 12 mm. The activities were measured by a well-calibrated high purity germanium detector. In the neutron source evaluation, since the triple-foil set was modeled in the MCNP, the raw reaction rates per atom were utilized without self-shielding correction. The two-dimensional 63Cu(n,g)64Cu reaction rate mapping was executed using the indirect neutron radiography. Its concept is similar to the foil activation technique; however, the radioactivity of the activation detector, which is a 15 20 0.0125 cm3 natural Cu sheet, is recorded via the imaging plate (IP) of type BAS-III manufactured by the Fujifilm. Since the IP response is proportional to the activity and therefore the reaction rate, the IP image can be translated into the two-dimensional 63Cu(n,g)64Cu reaction rate distribution, which is mostly induced by thermal neutrons (Liu et al., 2009c; Tsai et al., 2009). The paired ionization chamber method is commonly recommended to determine the dose and dose profile for mixed fields of neutrons and photons by solving a pair of algebraic equations as defined in ICRU 45 (1989) (see also Riley et al., 2004; Roca et al., 2009). One of the employed ionization chamber, denoted as TE(TE) chamber, is walled with A-150 tissue equivalent (TE) and flushed with TE gas; the other, denoted as Mg(Ar) chamber, is walled with magnesium and flushed with argon gas. The response of each chamber is normalized by its 60Co calibration factor and the neutron and gamma-ray dose components are evaluated with different depth-dependent sensitivities (Lin et al., 2010; Raaijmakers et al., 1995). Several corrections including ambient atmospheric temperature and pressure, ion recombination, wall thickness, and polarity of collecting potential have been considered in all measurements (Lin et al., this issue). 2.3. Experimental setup The evaluation work was performed in three beam conditions: without and with two different extension polyethylene (PE) collimators. The aperture of one collimator is 14 cm in diameter and the wall thickness is 4 cm; the aperture of the other is 8 cm in diameter and the wall thickness is 1 cm; both lengths are 18 cm. The evaluations were carried out in four different PMMA phantoms: two cubic phantoms with side lengths 15 cm and 21 cm, as well as two specially assembled PMMA phantoms. The special PMMA phantoms were assembled and applied for the sake of evaluating the neutron source in more complicated and critical geometries. Fig. 1 shows one of the special phantoms, which is a pile of square blocks with the same square area 20 20 cm2 and different heights from 2.88 to 5.24 cm. The other one is a stack of PMMA cylinders/disks, whose diameters are from 10.0 to 22.5 cm, and heights are from 1 to 4 cm; the model of this phantom is illustrated by Fig. 2.
Fig. 1. The special phantom assembled by PMMA blocks with different thicknesses.
Fig. 2. The special phantom assembled by PMMA cylinders/disks with different thicknesses and radii.
The evaluations of reaction rates and dose rates were executed at the depths of 1.5, 2.0, 2.5, 5.0, 8.0, 10.0, and 13.0 cm along the central axis inside the 15 15 15 cm3 phantom against the irradiation port aperture surface. Phantom central axis and beam central axis were coincident. The comparisons of reaction rate distributions along the central axis were carried out considering the 21 21 21 cm3 phantom in the following conditions: (a) phantom was against the opening beam exit, (b) phantom was against the exit of the 4 cm thick collimator, (c) phantom was against the exit of the 1 cm thick collimator, and (d) the front surface of the phantom was 10 cm away from the beam exit without any collimators. The reaction rate mappings along the central plane inside the specially assembled PMMA phantoms were compared as well. The distance from their vertical central axes to the opening beam exit is 15 cm for both cases. All positioning work was done by four cross-hair lasers. All the measurements performed in this work were done at a nominal reactor power of 1.2 MW. For the time-dependent beam intensity variation, all of the measured results were normalized to the reference state of 1.2 MW via a well calibrated on-line neutron monitoring system (Liu et al., this issue-b). The calculations were accordingly performed to obtain the corresponding reaction rates at the reference state.
1894
Y.-H. Liu et al. / Applied Radiation and Isotopes 69 (2011) 1892–1896
3. Results and discussions For the 21 21 21 cm3 PMMA phantom, Fig. 3 shows the measured distribution of the absolute 63Cu(n,g)64Cu reaction rate using the indirect neutron radiography, which has incorporated instrumental neutron activation analysis; it also presents the calculation in the according condition with measurement using MCNP with THOR-Y09, THOR-50C, and the surface source RSSA file. Note that the calculated values have a statistical uncertainty ranging from 0.5% up to 2% (at 15 cm) in 95% confidence interval; the measured values have experimental error of 2% due to the intrinsic uncertainty of IP mostly. Evidently, the source description THOR-Y09 works much better than THOR-50C with regard to the measured result. The large magnitude difference between the THOR-50C calculated and the IP measured reaction rates may be due to the imperfect core model used in the original calculation creating THOR-50C. Nevertheless, even after scaling the magnitude according to the maximum point of measured value, THOR-50C still overestimated the reaction rate before the peak ( 410%) and significantly underestimated afterwards. The improper reaction rate distribution, compared to THOR-Y09, was due to the lower resolution energy binning (47 groups vs. 640 groups), poorer spatial resolution
(1 cm width per ring vs. 1 mm width per ring), and the angular distribution structure (15 polar angular bins vs. 330 angular bins over the whole source surface). Comparing the measured result to the THOR-Y09 and RSSA results, although the RSSA and THOR-Y09 results are similar, THOR-Y09 is evidently closer to the measured curve (o5%). At the peak position ( 2.1 cm), the reaction rate difference between the calculation with THOR-Y09 and the measurement was smaller than 0.3%, while the deviation was 2.1% between the calculation with RSSA and the measurement. The relative 63Cu(n,g)64Cu reaction rate distributions along the central axis of the 21 21 21 cm3 cubic phantom, measured using the indirect neutron radiography and calculated using MCNP with THOR-Y09, are shown in Fig. 4 for different experimental configurations. All curves were normalized to their maximum value, 100%. For cases 1 and 2, the differences between the calculated value and the measured points were smaller than 2.5% before the maximum points, while the calculated values were about 3% higher than those from measurement in case 3. After the maximum points, the calculated curves turn to be few percent lower than measurement for all the three cases. The difference between the normalized, the calculated, and the measured 2D relative reaction rate distributions of the specially assembled phantom shown in Fig. 1 is between 3.65% and
Fig. 3. The measured and calculated (THOR-Y09, RSSA, and THOR-50C) absolute reaction rate distributions of the 63Cu(n,g)64Cu reaction in the 21 21 21 cm3 phantom.
Fig. 4. The measured and calculated relative 63Cu(n,g)64Cu reaction rate distributions along the central axis of the 21 21 21 cm3 PMMA phantom for (1) the case with 4 cm wall thickness, 18 cm long PE collimator, (2) the case with 1 cm wall thickness, 18 cm long PE collimator, and (3) the case without collimator at 10 cm away from the irradiation port aperture surface.
Y.-H. Liu et al. / Applied Radiation and Isotopes 69 (2011) 1892–1896
1895
Fig. 6. The measured and calculated absolute reaction rates of the triple activation foils, and the neutron and gamma-ray dose rates along the central axis of the 15 15 15 cm3 PMMA phantom.
gamma-ray dose rate, the deviations are lower than 5% as well. Regarding the total neutron dose rate, the calculated values are higher than measurement before the depth of 8 cm. Based on the good matches of calculated and measured reaction rates of the triple activation foils, the differences might be due to the large uncertainties of the derived neutron doses from the paired ionization chamber method, which may be higher than 30%.
4. Conclusions
Fig. 5. The difference maps between the measured and the MCNP calculated reaction rate distributions inside the special phantom, shown as Fig. 2, at the (a) 0 cm and (b) 4 cm height planes.
The comparisons performed in different phantoms for reaction rate distributions, absolute reaction rates of triple foils, and neutron and gamma-ray dose rates indicated that THOR-Y09 can work satisfactory to predict the reaction rate and dose rate in several different configurations. Therefore, THOR-Y09 will be used in the coming clinical trial. For the absolute reaction rates, the differences are all within 5% difference. For the 2D reaction rate distributions, most of the deviations are within 72.5%. Concerning the gamma-ray dose rate, the calculation resembles the measurement within 5% difference. As to the total neutron dose rate, the difference should be dominated by the measurement itself. In conclusion, THOR-Y09 is an appropriate source description not only for neutrons but also for gamma rays with regard to the configuration presented in this work and therefore to similar setups. In the future work, a customized MCNP5, with reference to the work of Albritton and Kiger, will be considered to improve the generalized source and tested for comparison.
Acknowledgments 1.45%. At most of the positions, the differences are within 72.5%. Fig. 5 presents the reaction rate difference maps of the other special phantom at the 0 and 4 cm height planes. At the 0 cm plane, the differences are between 3.1% and 1.45%; at the 4 cm plane, the differences are within 4.7% to 0.99%. If a tolerance of 5% deviation is chosen as a criterion, THOR-Y09 can deliver reliable results mostly within the above requirement. Note that the calculated values have a statistical uncertainty ranging from 0.5% up to 3.5% (at corner) in 95% confidence interval. Fig. 6 shows the reaction rates per atom of the AuAl, Cu, and MnNi foils, as well as the total neutron and gamma-ray dose rates measured in the 15 15 15 cm3 PMMA phantom at 7 different depths. The reaction rates calculated by MCNP with THOR-Y09 matched the measured values within 5% difference. For the
This work was supported by the National Science Council, R.O.C. (Taiwan) under contract number 99-2218-E-007-009. The authors would like to express their sincere appreciation to the Nuclear Reactor Division of the Nuclear Science and Technology Development Center, National Tsing Hua University, for their kind assistance in the reactor operation. References Albritton, J.R., Kiger III, W.S., 2008. Neutron beam source definition techniques for NCT treatment planning. In: Zonta, A. (Ed.), Proceedings of 13th ICNCT. ENEA, Rome, pp. 571–574. Albritton, J.R., Kiger III, W.S., 2010. Application of variance reduction techniques in Monte Carlo treatment planning calculations for NCT. In: Liberman (Ed.),
1896
Y.-H. Liu et al. / Applied Radiation and Isotopes 69 (2011) 1892–1896
Proceedings of 14th ICNCT. Comisio´n Nacional de Energı´a Ato´mica, Buenos Aires, Argentina, pp. 246–249. Chang, C.-H., Liu, Y.-H., Chen, A.-Y., Jiang, S.-H., Liu, H.Y.-W., Liu, H.-M., 2006. A study of neutron fluence rates of the BNCT Beam at THOR using foil activation. In: Nakagawa (Ed.), Proceedings of 12th ICNCT. International Society of Neutron Capture Therapy, Takamatsu, Kagawa, Japan, pp. 453–457. Raaijmakers, C.P., Konijnenberg, M.W., Verhagen, H.W., Mijnheer, B.J., 1995. Determination of dose components in phantoms irradiated with an epithermal neutron beam for boron neutron capture therapy. Medical Physics 22, 321–329. ICRU, 1989. Clinical neutron dosimetry part I: determination of absorbed dose in a patient treated by external beams of fast neutrons. ICRU Report 45. Lin, Y.-C., Roca, A., Liu, Y.-H., Tsai, P.-E., Nievaart, S., Liu, H.-M., Moss, R., Chou, W.T., Jiang, S.-H., 2010. The dose comparison between the THOR and HFR epithermal neutron beams. Radiation Measurements 45, 1422–1426. Lin, Y.-C., Liu, Y.-H., Jiang, S.-H., Liu, H.-M., Chou, W.-T., The refinement of dose assessment of the THOR BNCT beam. Applied Radiation and Isotopes, this issue. doi:10.1016/j.apradiso.2011.01.03. Liu, Y.-H., Nievaart, S., Tsai, P.-E., Liu, H.-M., Moss, R., Jiang, S.-H., 2009a. Neutron spectra measurement and comparison of the HFR and THOR BNCT beams. Applied Radiation and Isotopes 67, S137–S140. Liu, Y.-H., 2009b. Neutronic Characterization of an Epithermal Neutron Beam in Boron Neutron Capture Therapy. Doctoral Dissertation. National Tsing Hua University, Taiwan. Liu, Y.-H., Huang, C.-K., Tsai, P.-E., Chen, A.-Y., Liu, H.-M., Lee, S.-C., Jiang, S.-H., 2009c. BNCT epithermal neutron beam mapping by using indirect neutron radiography. Nuclear Technology 168, 354–358.
Liu, Y.-H., Tsai, P.-E., Liu, H.-M., Jiang, S.-H., 2010. The angular and spatial distributions of the thermal neutron source description of the THOR BNCT beam. Radiation Measurements 45, 1432–1435. Liu, Y.-H., Lin, Y.-C., Nievaart, S., Chou, W.-T., Liu, H.-M., Jiang, S.-H. Determination of the gamma-ray spectrum in a strong neutron/gamma-ray mixed field. Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, this issue-a. doi:10.1016/ j.nima.2010.11.184. Liu, Y.-H., Tsai, P.-E., Lin, Y.-C., Huang, C.-K., Liu, H.-M., Jiang, S.-H. Quality control and quality assurance procedures at the THOR BNCT facility. Applied Radiation and Isotopes, this issue-b. doi:10.1016/j.apradiso.2011.03.012. Riley, K.J., Binns, P.J., Harling, O.K., 2004. A state-of-the-art epithermal neutron irradiation facility for neutron capture therapy. Physics in Medicine and Biology 49, 3725. Roca, A., Liu, Y.-H., Wojnecki, C., Green, S., Nievaart, S., Ghani, Z., Moss, R., 2009. A preliminary inter-centre comparison study for photon, thermal neutron and epithermal neutron responses of two pairs of ionisation chambers used for BNCT. Applied Radiation and Isotopes 67, S134–S136. Tsai, P.-E., Liu, Y.-H., Huang, C.-K., Liu, H.-M., Jiang, S.-H., 2009. Neutron flux mapping inside a cubic and a head PMMA phantom using indirect neutron radiography. Applied Radiation and Isotopes 67, S190–S194. Tsai, P.-E., Liu, Y.-H., Liu, H.-M., Jiang, S.-H., 2010. Characterization of a BNCT beam using neutron activation and indirect neutron radiography. Radiation Measurements 45, 1167–1170. X-5 Monte Carlo Team, 2005. MCNP—A General Monte Carlo N-Particle Transport Code Version 5, LA-UR-03-1987, LA-CP-03-0245, and LA-CP-03-0284. Los Alamos National Laboratory, University of California.
Contents lists available at ScienceDirect
Applied Radiation and Isotopes journal homepage: www.elsevier.com/locate/apradiso
Performance evaluation of the source description of the THOR BNCT epithermal neutron beam Yuan-Hao Liu a,n, Pi-En Tsai a, Hui-Ting Yu b, Yi-Chun Lin c, Yu-Shiang Huang b, Chun-Kai Huang b, Yen-Wan Hsueh Liu b, Hong-Ming Liu a, Shiang-Huei Jiang b a
Nuclear Science and Technology Development Center, National Tsing Hua University, Hsinchu, Taiwan Institute of Nuclear Engineering and Science, National Tsing Hua University, HsinChu, Taiwan c Department of Biomedical Engineering and Environmental Sciences, National Tsing Hua University, HsinChu, Taiwan b
a r t i c l e i n f o
a b s t r a c t
Available online 13 April 2011
This paper aims to evaluate the performance of the source description of the THOR BNCT beam via different measurement techniques in different phantoms. The measurement included (1) the absolute reaction rate measurement of a set of triple activation foils, (2) the neutron and gamma-ray dose rates measured using the paired ionization chamber method, and (3) the relative reaction rate distributions obtained using the indirect neutron radiography. Three source descriptions, THOR-Y09, surface source file RSSA, and THOR-50C, were tested. The comparison results concluded that THOR-Y09 is a welltested source description not only for neutron components, but also for gamma-ray component. & 2011 Elsevier Ltd. All rights reserved.
Keywords: THOR BNCT Epithermal neutron beam Source description
1. Introduction In 2004, the Tsing Hua Open-pool Reactor (THOR) in Hsinchu, a 2 MW TRIGA reactor, was modified into a boron neutron capture therapy (BNCT) facility with an intense epithermal neutron beam, reaching 1.07 109 epithermal neutrons cm 2 s 1 at 1.2 MW (Liu et al., 2009a). For BNCT, accurately predicting dose delivery depends not only on determination of boron distribution and concentration but also on validness of the treatment planning system. Hence, a source description plays a very important role in treatment planning system; it is necessary to validate the implemented source description of the used BNCT beam to ensure the correctness of calculated dose. Correspondingly, it is of our interest to evaluate the performance of the THOR source description in different scenarios. To achieve such a purpose, a series of delicate beam dosimetry, characterization, and adjustment have been performed; activation detectors, indirect neutron radiography, and paired ionization chambers were employed as measuring tools. Measurements were conducted in different PMMA phantoms. The following sections will detail the evaluation work.
2. Materials and methods 2.1. Source description Generally, a source description consists of neutron and gamma-ray components. At THOR, the generalized source n
Corresponding author. Tel.: þ886 3 5715131x42889. E-mail addresses: [email protected], [email protected] (Y.-H. Liu).
0969-8043/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2011.04.005
description, denoted as THOR-Y09, includes four sets of sources: fast neutron, epithermal neutron, thermal neutron, and gamma-ray sub-sources. Each sub-source has its own energy, angular, and spatial distributions. THOR-Y09 was an experimentally adjusted version of the calculated, generalized source description THOR-50C, which has only two sets of sources: one neutron and one gamma-ray sub-sources. The neutron source adjustment utilized full-scale deconvolution; this innovative methodology not only deconvoluted the energy distribution but also adjusted the angular and spatial distributions (Liu, 2009b; Liu et al., 2010). The gamma-ray source was adjusted through a multi-cap magnesium-argon ionization chamber with the aid of Monte Carlo calculated detector response functions (Liu et al., this issue-a). The source surfaces of THOR-50C and THOR-Y09 are both located at the irradiation port aperture surface. In addition to THOR-Y09 and THOR-50C, a surface source RSSA was created using the Monte Carlo code MCNP5 (X-5 Monte Carlo Team, 2005) with a relatively modified/improved model of reactor core and beam configuration compared to the one used in generating THOR-50C. Surface source distribution has a significant advantage that its azimuthal angle information is more faithful than generalized source description (Albritton and Kiger, 2008). However, due to the huge amount of data it records to achieve trustiness, the RSSA is not handy to use. For example, in our test, RSSA file cannot be used via MPI interface to make parallel calculation; the applied RSSA file has a size of 60 GB. Furthermore, when a discrepancy is found between measurement and calculation, it is important that the source is adjustable, which is unavailable for the RSSA file. Unfortunately, this is the case at THOR, despite the fact that with proper variance reduction
Y.-H. Liu et al. / Applied Radiation and Isotopes 69 (2011) 1892–1896
1893
techniques the size of RSSA could be effectively reduced (Albritton and Kiger, 2010). Nevertheless, using a customized version of MCNP5 such as the one modified by Albritton and Kiger (2008), the RSSA file can be used as a set of default parameters built outside MCNP. Until present, however, the modified version is not yet available at THOR but will be considered in the future work. No matter how delicate a source description is, it has to be proven reliable and trustworthy. To prove the trustiness of THOR-Y09, tests were carried out in this work. 2.2. Measurement tools The triple-foil set plays a role as activation detectors to measure the neutron intensity in the study (Chang et al., 2006; Tsai et al., 2010). It consists of AuAl (1 wt% Au, 0.2 mm thick), natural Cu (0.1 mm thick), and MnNi (88 wt% Mn, 0.1 mm thick) foils in sequence. Foil diameter is 12 mm. The activities were measured by a well-calibrated high purity germanium detector. In the neutron source evaluation, since the triple-foil set was modeled in the MCNP, the raw reaction rates per atom were utilized without self-shielding correction. The two-dimensional 63Cu(n,g)64Cu reaction rate mapping was executed using the indirect neutron radiography. Its concept is similar to the foil activation technique; however, the radioactivity of the activation detector, which is a 15 20 0.0125 cm3 natural Cu sheet, is recorded via the imaging plate (IP) of type BAS-III manufactured by the Fujifilm. Since the IP response is proportional to the activity and therefore the reaction rate, the IP image can be translated into the two-dimensional 63Cu(n,g)64Cu reaction rate distribution, which is mostly induced by thermal neutrons (Liu et al., 2009c; Tsai et al., 2009). The paired ionization chamber method is commonly recommended to determine the dose and dose profile for mixed fields of neutrons and photons by solving a pair of algebraic equations as defined in ICRU 45 (1989) (see also Riley et al., 2004; Roca et al., 2009). One of the employed ionization chamber, denoted as TE(TE) chamber, is walled with A-150 tissue equivalent (TE) and flushed with TE gas; the other, denoted as Mg(Ar) chamber, is walled with magnesium and flushed with argon gas. The response of each chamber is normalized by its 60Co calibration factor and the neutron and gamma-ray dose components are evaluated with different depth-dependent sensitivities (Lin et al., 2010; Raaijmakers et al., 1995). Several corrections including ambient atmospheric temperature and pressure, ion recombination, wall thickness, and polarity of collecting potential have been considered in all measurements (Lin et al., this issue). 2.3. Experimental setup The evaluation work was performed in three beam conditions: without and with two different extension polyethylene (PE) collimators. The aperture of one collimator is 14 cm in diameter and the wall thickness is 4 cm; the aperture of the other is 8 cm in diameter and the wall thickness is 1 cm; both lengths are 18 cm. The evaluations were carried out in four different PMMA phantoms: two cubic phantoms with side lengths 15 cm and 21 cm, as well as two specially assembled PMMA phantoms. The special PMMA phantoms were assembled and applied for the sake of evaluating the neutron source in more complicated and critical geometries. Fig. 1 shows one of the special phantoms, which is a pile of square blocks with the same square area 20 20 cm2 and different heights from 2.88 to 5.24 cm. The other one is a stack of PMMA cylinders/disks, whose diameters are from 10.0 to 22.5 cm, and heights are from 1 to 4 cm; the model of this phantom is illustrated by Fig. 2.
Fig. 1. The special phantom assembled by PMMA blocks with different thicknesses.
Fig. 2. The special phantom assembled by PMMA cylinders/disks with different thicknesses and radii.
The evaluations of reaction rates and dose rates were executed at the depths of 1.5, 2.0, 2.5, 5.0, 8.0, 10.0, and 13.0 cm along the central axis inside the 15 15 15 cm3 phantom against the irradiation port aperture surface. Phantom central axis and beam central axis were coincident. The comparisons of reaction rate distributions along the central axis were carried out considering the 21 21 21 cm3 phantom in the following conditions: (a) phantom was against the opening beam exit, (b) phantom was against the exit of the 4 cm thick collimator, (c) phantom was against the exit of the 1 cm thick collimator, and (d) the front surface of the phantom was 10 cm away from the beam exit without any collimators. The reaction rate mappings along the central plane inside the specially assembled PMMA phantoms were compared as well. The distance from their vertical central axes to the opening beam exit is 15 cm for both cases. All positioning work was done by four cross-hair lasers. All the measurements performed in this work were done at a nominal reactor power of 1.2 MW. For the time-dependent beam intensity variation, all of the measured results were normalized to the reference state of 1.2 MW via a well calibrated on-line neutron monitoring system (Liu et al., this issue-b). The calculations were accordingly performed to obtain the corresponding reaction rates at the reference state.
1894
Y.-H. Liu et al. / Applied Radiation and Isotopes 69 (2011) 1892–1896
3. Results and discussions For the 21 21 21 cm3 PMMA phantom, Fig. 3 shows the measured distribution of the absolute 63Cu(n,g)64Cu reaction rate using the indirect neutron radiography, which has incorporated instrumental neutron activation analysis; it also presents the calculation in the according condition with measurement using MCNP with THOR-Y09, THOR-50C, and the surface source RSSA file. Note that the calculated values have a statistical uncertainty ranging from 0.5% up to 2% (at 15 cm) in 95% confidence interval; the measured values have experimental error of 2% due to the intrinsic uncertainty of IP mostly. Evidently, the source description THOR-Y09 works much better than THOR-50C with regard to the measured result. The large magnitude difference between the THOR-50C calculated and the IP measured reaction rates may be due to the imperfect core model used in the original calculation creating THOR-50C. Nevertheless, even after scaling the magnitude according to the maximum point of measured value, THOR-50C still overestimated the reaction rate before the peak ( 410%) and significantly underestimated afterwards. The improper reaction rate distribution, compared to THOR-Y09, was due to the lower resolution energy binning (47 groups vs. 640 groups), poorer spatial resolution
(1 cm width per ring vs. 1 mm width per ring), and the angular distribution structure (15 polar angular bins vs. 330 angular bins over the whole source surface). Comparing the measured result to the THOR-Y09 and RSSA results, although the RSSA and THOR-Y09 results are similar, THOR-Y09 is evidently closer to the measured curve (o5%). At the peak position ( 2.1 cm), the reaction rate difference between the calculation with THOR-Y09 and the measurement was smaller than 0.3%, while the deviation was 2.1% between the calculation with RSSA and the measurement. The relative 63Cu(n,g)64Cu reaction rate distributions along the central axis of the 21 21 21 cm3 cubic phantom, measured using the indirect neutron radiography and calculated using MCNP with THOR-Y09, are shown in Fig. 4 for different experimental configurations. All curves were normalized to their maximum value, 100%. For cases 1 and 2, the differences between the calculated value and the measured points were smaller than 2.5% before the maximum points, while the calculated values were about 3% higher than those from measurement in case 3. After the maximum points, the calculated curves turn to be few percent lower than measurement for all the three cases. The difference between the normalized, the calculated, and the measured 2D relative reaction rate distributions of the specially assembled phantom shown in Fig. 1 is between 3.65% and
Fig. 3. The measured and calculated (THOR-Y09, RSSA, and THOR-50C) absolute reaction rate distributions of the 63Cu(n,g)64Cu reaction in the 21 21 21 cm3 phantom.
Fig. 4. The measured and calculated relative 63Cu(n,g)64Cu reaction rate distributions along the central axis of the 21 21 21 cm3 PMMA phantom for (1) the case with 4 cm wall thickness, 18 cm long PE collimator, (2) the case with 1 cm wall thickness, 18 cm long PE collimator, and (3) the case without collimator at 10 cm away from the irradiation port aperture surface.
Y.-H. Liu et al. / Applied Radiation and Isotopes 69 (2011) 1892–1896
1895
Fig. 6. The measured and calculated absolute reaction rates of the triple activation foils, and the neutron and gamma-ray dose rates along the central axis of the 15 15 15 cm3 PMMA phantom.
gamma-ray dose rate, the deviations are lower than 5% as well. Regarding the total neutron dose rate, the calculated values are higher than measurement before the depth of 8 cm. Based on the good matches of calculated and measured reaction rates of the triple activation foils, the differences might be due to the large uncertainties of the derived neutron doses from the paired ionization chamber method, which may be higher than 30%.
4. Conclusions
Fig. 5. The difference maps between the measured and the MCNP calculated reaction rate distributions inside the special phantom, shown as Fig. 2, at the (a) 0 cm and (b) 4 cm height planes.
The comparisons performed in different phantoms for reaction rate distributions, absolute reaction rates of triple foils, and neutron and gamma-ray dose rates indicated that THOR-Y09 can work satisfactory to predict the reaction rate and dose rate in several different configurations. Therefore, THOR-Y09 will be used in the coming clinical trial. For the absolute reaction rates, the differences are all within 5% difference. For the 2D reaction rate distributions, most of the deviations are within 72.5%. Concerning the gamma-ray dose rate, the calculation resembles the measurement within 5% difference. As to the total neutron dose rate, the difference should be dominated by the measurement itself. In conclusion, THOR-Y09 is an appropriate source description not only for neutrons but also for gamma rays with regard to the configuration presented in this work and therefore to similar setups. In the future work, a customized MCNP5, with reference to the work of Albritton and Kiger, will be considered to improve the generalized source and tested for comparison.
Acknowledgments 1.45%. At most of the positions, the differences are within 72.5%. Fig. 5 presents the reaction rate difference maps of the other special phantom at the 0 and 4 cm height planes. At the 0 cm plane, the differences are between 3.1% and 1.45%; at the 4 cm plane, the differences are within 4.7% to 0.99%. If a tolerance of 5% deviation is chosen as a criterion, THOR-Y09 can deliver reliable results mostly within the above requirement. Note that the calculated values have a statistical uncertainty ranging from 0.5% up to 3.5% (at corner) in 95% confidence interval. Fig. 6 shows the reaction rates per atom of the AuAl, Cu, and MnNi foils, as well as the total neutron and gamma-ray dose rates measured in the 15 15 15 cm3 PMMA phantom at 7 different depths. The reaction rates calculated by MCNP with THOR-Y09 matched the measured values within 5% difference. For the
This work was supported by the National Science Council, R.O.C. (Taiwan) under contract number 99-2218-E-007-009. The authors would like to express their sincere appreciation to the Nuclear Reactor Division of the Nuclear Science and Technology Development Center, National Tsing Hua University, for their kind assistance in the reactor operation. References Albritton, J.R., Kiger III, W.S., 2008. Neutron beam source definition techniques for NCT treatment planning. In: Zonta, A. (Ed.), Proceedings of 13th ICNCT. ENEA, Rome, pp. 571–574. Albritton, J.R., Kiger III, W.S., 2010. Application of variance reduction techniques in Monte Carlo treatment planning calculations for NCT. In: Liberman (Ed.),
1896
Y.-H. Liu et al. / Applied Radiation and Isotopes 69 (2011) 1892–1896
Proceedings of 14th ICNCT. Comisio´n Nacional de Energı´a Ato´mica, Buenos Aires, Argentina, pp. 246–249. Chang, C.-H., Liu, Y.-H., Chen, A.-Y., Jiang, S.-H., Liu, H.Y.-W., Liu, H.-M., 2006. A study of neutron fluence rates of the BNCT Beam at THOR using foil activation. In: Nakagawa (Ed.), Proceedings of 12th ICNCT. International Society of Neutron Capture Therapy, Takamatsu, Kagawa, Japan, pp. 453–457. Raaijmakers, C.P., Konijnenberg, M.W., Verhagen, H.W., Mijnheer, B.J., 1995. Determination of dose components in phantoms irradiated with an epithermal neutron beam for boron neutron capture therapy. Medical Physics 22, 321–329. ICRU, 1989. Clinical neutron dosimetry part I: determination of absorbed dose in a patient treated by external beams of fast neutrons. ICRU Report 45. Lin, Y.-C., Roca, A., Liu, Y.-H., Tsai, P.-E., Nievaart, S., Liu, H.-M., Moss, R., Chou, W.T., Jiang, S.-H., 2010. The dose comparison between the THOR and HFR epithermal neutron beams. Radiation Measurements 45, 1422–1426. Lin, Y.-C., Liu, Y.-H., Jiang, S.-H., Liu, H.-M., Chou, W.-T., The refinement of dose assessment of the THOR BNCT beam. Applied Radiation and Isotopes, this issue. doi:10.1016/j.apradiso.2011.01.03. Liu, Y.-H., Nievaart, S., Tsai, P.-E., Liu, H.-M., Moss, R., Jiang, S.-H., 2009a. Neutron spectra measurement and comparison of the HFR and THOR BNCT beams. Applied Radiation and Isotopes 67, S137–S140. Liu, Y.-H., 2009b. Neutronic Characterization of an Epithermal Neutron Beam in Boron Neutron Capture Therapy. Doctoral Dissertation. National Tsing Hua University, Taiwan. Liu, Y.-H., Huang, C.-K., Tsai, P.-E., Chen, A.-Y., Liu, H.-M., Lee, S.-C., Jiang, S.-H., 2009c. BNCT epithermal neutron beam mapping by using indirect neutron radiography. Nuclear Technology 168, 354–358.
Liu, Y.-H., Tsai, P.-E., Liu, H.-M., Jiang, S.-H., 2010. The angular and spatial distributions of the thermal neutron source description of the THOR BNCT beam. Radiation Measurements 45, 1432–1435. Liu, Y.-H., Lin, Y.-C., Nievaart, S., Chou, W.-T., Liu, H.-M., Jiang, S.-H. Determination of the gamma-ray spectrum in a strong neutron/gamma-ray mixed field. Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, this issue-a. doi:10.1016/ j.nima.2010.11.184. Liu, Y.-H., Tsai, P.-E., Lin, Y.-C., Huang, C.-K., Liu, H.-M., Jiang, S.-H. Quality control and quality assurance procedures at the THOR BNCT facility. Applied Radiation and Isotopes, this issue-b. doi:10.1016/j.apradiso.2011.03.012. Riley, K.J., Binns, P.J., Harling, O.K., 2004. A state-of-the-art epithermal neutron irradiation facility for neutron capture therapy. Physics in Medicine and Biology 49, 3725. Roca, A., Liu, Y.-H., Wojnecki, C., Green, S., Nievaart, S., Ghani, Z., Moss, R., 2009. A preliminary inter-centre comparison study for photon, thermal neutron and epithermal neutron responses of two pairs of ionisation chambers used for BNCT. Applied Radiation and Isotopes 67, S134–S136. Tsai, P.-E., Liu, Y.-H., Huang, C.-K., Liu, H.-M., Jiang, S.-H., 2009. Neutron flux mapping inside a cubic and a head PMMA phantom using indirect neutron radiography. Applied Radiation and Isotopes 67, S190–S194. Tsai, P.-E., Liu, Y.-H., Liu, H.-M., Jiang, S.-H., 2010. Characterization of a BNCT beam using neutron activation and indirect neutron radiography. Radiation Measurements 45, 1167–1170. X-5 Monte Carlo Team, 2005. MCNP—A General Monte Carlo N-Particle Transport Code Version 5, LA-UR-03-1987, LA-CP-03-0245, and LA-CP-03-0284. Los Alamos National Laboratory, University of California.