Home
Search
Collections
Journals
About
Contact us
My IOPscience
Monte Carlo simulations on performance of double-scattering Compton camera
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2012 JINST 7 C01009 (http://iopscience.iop.org/1748-0221/7/01/C01009) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 221.130.23.86 This content was downloaded on 19/10/2013 at 18:53
Please note that terms and conditions apply.
P UBLISHED BY IOP P UBLISHING FOR SISSA R ECEIVED: December 7, 2011 ACCEPTED: December 7, 2011 P UBLISHED: January 3, 2012
13th I NTERNATIONAL W ORKSHOP 3–7 J ULY 2011, ETH Z URICH , S WITZERLAND
ON
R ADIATION I MAGING D ETECTORS ,
J.H. Park,a H. Seo,a Y.S. Kim,a C.H. Kim,a,1 J.H. Lee,b C.S. Lee,b S.M. Kimc and J.S. Leec a Department
of Nuclear Engineering, Hanyang University, 17 Haengdang, Seongdong, Seoul 133-791, Korea b Department of Physics, Chung-Ang University, 221 Heukseok, Dongjak, Seoul 156-756, Korea c Department of Nuclear Medicine, Seoul National University, 28 Yeongeon, Jongno, Seoul 110-744, Korea
E-mail:
[email protected] A BSTRACT: A double-scattering Compton camera that can effectively obtain three-dimensional emission images in high-energy gamma-ray applications such as nuclear decommissioning and particle therapy monitoring has been developed. The double-scattering Compton camera utilizes two position-sensitive detectors as scatter detectors to determine the trajectory of a scattered gammaray, and a scintillation detector as absorber detector to measure the remaining energy of the doublescattered gamma-ray. The benefit of using two scatter detectors is the accurate determination of the gamma-ray trajectory after the scattering at the first scatter detector, which makes possible higher imaging resolution. In the present study, Geant4 Monte Carlo simulations were conducted to compare the performance of the double-scattering Compton camera with that of a similarly dimensioned single-scattering Compton camera for different source energies. Further, the optimal geometry of the multiple-detector-type double-scattering Compton camera was investigated for the purposes of increasing its imaging sensitivity. The results showed that the double-scattering Compton camera offers superior angular resolution over the entire energy range considered in the present study, whereas the single-scattering Compton camera provides greater sensitivity. The results also showed that, in general, the placement of additional detectors in the axial direction (i.e., stacking) is more effective for sensitivity improvement than doing so in the planar direction. This axial placement, however, lowers the imaging resolution. The double-scattering Compton camera exhibited the highest sensitivity when the additional scatter detectors were added to the first scatter detector 1 Corresponding
author.
c 2012 IOP Publishing Ltd and SISSA
doi:10.1088/1748-0221/7/01/C01009
2012 JINST 7 C01009
Monte Carlo simulations on performance of double-scattering Compton camera
in the axial direction, and exhibited the highest imaging resolution when the additional detectors were added to first scatter detector in the planar direction. Therefore, the results generally indicated that the first scatter detector is more important than the second not only for improving the sensitivity but also for maintaining a high imaging resolution. We believe that the present study’s findings will provide valuable guidelines to researchers looking for the best Compton camera designs for certain objectives. K EYWORDS : Models and simulations; Compton imaging
2012 JINST 7 C01009
Contents Double-scattering Compton camera
1
2
Simulation geometry
2
3
Comparison with single-scattering Compton camera
3
4
Multiple-detector-type double-scattering Compton camera
4
5
Conclusion
5
1
Double-scattering Compton camera
Compared with conventional gamma-ray emission imaging devices based on mechanical collimation, the Compton imaging technique, based on the principle known as “electronic collimation,” could provide better performance in terms of imaging resolution and sensitivity when the gammaray source to be imaged has relatively high energy (i.e., more than a few hundred keV) [1, 2]. The imaging resolution of the Compton camera depends on the accuracy of the reconstructed cone for an incident gamma-ray, which accuracy is affected by several detector parameters including the energy and spatial resolution of the component detectors. However, a typical Compton camera, consisting of one scatter detector and one absorber detector [3], has a limitation on its absorber detector’s spatial resolution, especially in high-energy gamma-ray applications, because the absorber detector’s pixel size has to be kept large enough to fully absorb a scattered gamma-ray within a single pixel, and it consequently degrades the imaging resolution. Our research group has developed, as a simple solution to this problem, a prototype double-scattering Compton camera consisting of two double-sided silicon strip detectors (DSSDs) as scatter detectors and a cylindrical NaI(Tl) scintillation detector as the absorber detector. This type Compton camera is suitable for either nuclear decommissioning applications or particle therapy applications, both of which involve imaging of high-energy gamma-ray sources [4, 5]. The basic idea of the double-scattering Compton camera is the maximization of imaging resolution by very accurate measurement of two successive Compton-scattering interaction positions using two thin, high-spatial-resolution scatter detectors followed by measurement of the remaining energy of the double-scattered gamma-ray using an absorber detector. As shown in the left side of figure 1, if a gamma-ray is scattered in each scatter detector sequentially and completely absorbed in the absorber detector, we can get a cone from the measured information of the interaction positions and deposited energies in the three component detectors. The interaction position in the first scatter detector (x1 , y1 , z1 ) becomes the apex of the backprojection cone. The axis of the cone is determined from the two interaction positions, i.e., (x1 , y1 , z1 ) and (x2 , y2 , z2 ). The opening angle of the cone is calculated from the measured energies and the equation (1.1). The superimposition
–1–
2012 JINST 7 C01009
1
of these conical surfaces in the three-dimensional image space gives us a reconstructed Compton image [6]. 1 1 − (1.1) θ = cos−1 1 + mo c2 E0 E0 − E1 In the present study, Geant4 Monte Carlo simulations [7, 8] were conducted to compare the performance of a prototype double-scattering Compton camera with that of a similarly dimensioned single-scattering Compton camera for different source energies. Additionally, for higher imaging sensitivity, the optimal geometry of the multiple-detector-type double-scattering Compton camera was investigated by comparing its performance in two extreme cases: the placement of three additional scatter detectors in the planar and axial directions, respectively.
2
Simulation geometry
The prototype double-scattering Compton camera considered here consists of three detectors: two DSSDs (5 cm×5 cm×0.15 cm, 16×16 strips) and one cylindrical NaI(Tl) scintillation detector (3 in. (H)×3 in. (D)) [4]. The distance between the first and the second scatter detector was set as 10 cm, and the absorber detector was placed 1 cm behind the second scatter detector [9]. The performance of the Compton camera was simulated realistically including the detector energy resolutions, interaction position resolutions, and energy discrimination levels of the component detectors. Doppler energy broadening also was included in the simulations, using Geant4’s Livermore physics model. To reduce the computation time, the angle biasing technique, which limits the primary particles’ angle to the Compton camera, was used, and the particle weight was reduced accordingly. The performance of the Compton camera was evaluated with respect to two quantitative factors, which are sensitivity and angular resolution. The sensitivity was defined as the number of effective events divided by the number of gamma-rays emitted from the source; the angular resolution, meanwhile, was determined from the angular resolution measure (ARM), which is defined as the angle between the reconstructed Compton cone and the actual source direction [10]. Further, the figure-of-merit (FOM) was applied in order to consider the sensitivity and angular resolution simultaneously [11].
–2–
2012 JINST 7 C01009
Figure 1. Schematic diagram of prototype double-scattering Compton camera showing back-projected cone for single gamma-ray (left) and photograph of constructed imaging system (right).
3
Comparison with single-scattering Compton camera
Considering that the interaction cross section of photoelectric effect rapidly decreases with increasing source energy, showing approximately hv−m dependency (with m ≈ 3 at hv = 0.1 MeV and m ≈ 1 at hv = 5 MeV) [12], whereas the interaction cross section of Compton scattering decreases relatively slowly, it is expected that Compton scattering will dominate in the second detector over the photoelectric effect if the source energy is high enough even though the photon will lose some of its energy by the Compton scattering at the first detector. In the present study, the performance of the double-scattering Compton camera was compared with that of a similarly dimensioned singlescattering Compton camera developed at Chung-Ang University in Korea [13]. This latter Compton camera consists of two position-sensitive detectors: a DSSD (5 cm×5 cm×0.15 cm, 16 strips on each side) as the scatter detector and a 25-segmented germanium detector (5 cm×5 cm×2 cm, 5×5 pixels) as the absorber detector. The distance between the two detectors was also set as 10 cm. The performance of the two Compton camera types was compared for a point source positioned 5 cm in front of the camera for gamma-ray energies of 364, 511, 662, 1332, and 2000 keV. Figure 2 shows the evaluated sensitivities, angular resolutions, and FOMs of the doublescattering and single-scattering Compton cameras as functions of the source energy. As regards sensitivity, the single-scattering Compton camera showed superior performance over the entire energy range, though the difference tended to narrow as the source energy increased, due to the fact that the number of multi-pixel events in the absorber detector, which were not included among the effective events, increased as the source energy increased, thereby significantly diminishing the sensitivity of the single-scattering Compton camera. In the case of the double-scattering Compton camera, multi-pixel events hardly occurred in the second scatter detector, given that it is very thin (=1.5 mm), and also because the absorber detector of the double-scattering Compton camera is not position-sensitive, being used only to determine the energy of a double-scattered gamma-ray. In both types of Compton camera, as the source energy increased, the angular resolution improved; this, an intrinsic characteristic of any Compton camera, is owed mainly to the decrease of the Doppler energy broadening effect [10] and the increase of the energy resolution. Over the entire energy range, the double-scattering Compton camera showed higher imaging resolution than did the single-scattering type, a result that can be ascribed to the former’s use of two thin scatter detectors for higher-accuracy determination of the scattered gamma-ray track. The single-scattering Compton camera, with its higher sensitivity, showed better FOM performance, though significantly,
–3–
2012 JINST 7 C01009
Figure 2. Variation of sensitivity, angular resolution, and FOM for double-scattering Compton camera (black squares) and single-scattering Compton camera (red circles) as function of source energy.
the difference grew smaller as the source energy increased; indeed, at the source energy of 2 MeV, the double-scattering Compton camera took the lead in this regard. These results indicated that, when high sensitivity is required, the single-scattering type is the better choice, but that when high imaging resolution is crucial, or when the source to be imaged has an energy over 2 MeV — for example, prompt gamma measurement in particle therapy applications — the double-scattering type takes precedence.
4
Multiple-detector-type double-scattering Compton camera
As discussed in the previous section, the double-scattering Compton camera has a relatively low sensitivity because it uses very thin detectors for the scatter detectors; hence, the use of extra, multiple scatter detectors would seem to be a rational approach to its improvement. Basically, there are two simple ways to improve the imaging sensitivities: (1) by increasng the effective thickness of the scatter detectors by using multiple detectors in the axial diretion, and (2) by increaseing the effective detection area of the scatter detectors by using multiple detectors in the planar direction. To that end, the performance of the multiple-detector-type double-scattering Compton camera was evaluated for two extreme cases: (1) the addition of three, 1 cm-separated scatter detectors to the first or second scatter detector in the axial direction to increase an effective detector thickness, and (2) the same for the planar direction to increase an effective detection area (figure 3). The performances were compared for the sensitivity, angular resolution, and FOM. It was found that, in general, the placement of detectors in the axial direction (i.e., stacking) was more effective for sensitivity improvement than that in the planar direction, for both the first and second detectors. The results indicated that one can obtain a better performance enhancing the interac-
–4–
2012 JINST 7 C01009
Figure 3. Simulation geometries of multiple-detector-type double-scattering Compton camera in Open Inventor view provided by Geant4.
Table 1. Performances according to geometry of multiple-detector-type double-scattering Compton cameras for different source energies. As pertains to the type, “SC1 wide” references the planar-direction placement to the first scatter detector and “SC2 stack,” the axial-direction placement to the second scatter detector. Source energy Type
Sensitivity (×10−7 ) 12.1±0.60 12.7±1.16 22.9±1.01 17.4±0.31
Angular resolution [◦ ] 3.80±0.1 4.14±0.01 4.21±0.03 4.72±0.1
1332 keV FOM 22.03±1.09 18.02±1.65 30.66±1.35 16.68±0.29
Sensitivity (×10−7 ) 7.24±0.76 8.25±0.50 8.50±0.76 7.25±0.62
Angular resolution [◦ ] 3.68±0.04 3.82±0.03 4.02±0.05 4.50±0.09
FOM 14.5±1.52 14.7±0.89 13.04±1.16 7.83±0.67
tion probability by increasing the detector thickness rather than the detection area. The maximum difference, for the first scatter detectors and the lower energy source (662 keV), was a factor of about 2. Considering that the sensitivity of the prototype double-scattering Compton camera is 6.17×10−7 and 3.86×10−7 for 662 keV and 1332 keV (see figure 2), table 1 showed that this was improved by up to 3.71 and 2.20 times for the source energies of 662 and 1332 keV, respectively, when the additional scatter detectors were placed on the first scatter plane in the axial direction. However, this axial-direction solution resulted in lower imaging resolution than does the planardirection alternative. This is mainly because the distance between the first and second interaction positions decreases when the additional detectors are placed in the axial direction which increases the error in the axis of the backprojection cone. In addition to this effect, if the area of the scatter detector increases, a scattered photon with a larger scattering angle has a higher chance to become an effective event. The large scattering angle means the large amount of energy deposition in the detector; hence, the energy resolution is improved, resulting in the improved imaging resolution. The double-scattering Compton camera showed the highest sensitivity when the extra scatter detectors were added to the first scatter detector in the axial direction, and shows the highest imaging resolution when the extra detectors are added to the first scatter detector in the planar direction. The results generally indicated that the first scatter detector is more important than the second, not only for improving the sensitivity but also for maintaining a high imaging resolution.
5
Conclusion
In the present study, the performance of the double-scattering Compton camera was compared with a single-scattering Compton camera for different source energies (364, 511, 662, 1332, and 2000 keV). According to our simulation results, the double-scattering Compton camera shows superior angular resolution over the entire energy range, whereas the single-scattering type offers higher sensitivity. Note that the double-scattering type would be the reasonable choice for highenergy (i.e., more than 2 MeV) gamma-ray applications or in cases where high imaging resolution is crucial. When improving the sensitivity of the double-scattering Compton camera by means of extra scatter detectors, their addition to the first scatter detector in the axial direction provides the maximum benefit. We believe that this simulation study’s findings will provide valuable guidelines to researchers looking for the best Compton camera designs for certain objectives.
–5–
2012 JINST 7 C01009
SC1 wide SC2 wide SC1 stack SC2 stack
662 keV
Acknowledgments This research was supported by National Nuclear R&D Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (No. 2010-0023825, 2010-0028913, 2010-0018572).
References [1] R.W. Todd et al., A proposed gamma camera, Nature 251 (1974) 132.
[3] S. Motomura et al., Multiple molecular simultaneous imaging in a live mouse using semiconductor Compton camera, J. Anal. Atom. Spectrom. 23 (2008) 1089. [4] H. Seo et al., Development of double-scattering-type Compton camera with double-sided silicon strip detectors and NaI(Tl) scintillation detector, Nucl. Instrum. Meth. A 615 (2010) 333. [5] H. Seo et al., Multitracing capability of double-scattering Compton imager with NaI(Tl) scintillator absorber, IEEE Trans. Nucl. Sci. 57 (2010) 1420. [6] M. Singh and D. Doria, An electronically collimated gamma camera for single photon emission computed tomography. part II: Image reconstruction and preliminary experimental measurements, Med. Phys. 10 (1983) 428. [7] S. Agostinelli et al., GEANT4—A simulation toolkit, Nucl. Instrum. Meth. A 506 (2003) 250. [8] J. Allison et al., GEANT4 developments and applications, IEEE Trans. Nucl. Sci. 53 (2006) 270. [9] H. Seo et al., Optimal geometrical configuration of a double-scattering Compton camera for maximum imaging resolution and sensitivity, Nucl. Instrum. Meth. A 591 (2008) 80. [10] A. Zoglauer et al., Doppler broadening as a lower limit to the angular resolution of next generation Compton telescopes, Proc. SPIE 4851 (2003) 1302. [11] S.H. Lee et al., Performance evaluation of a table-top Compton camera for various detector parameters, Nucl. Instrum. Meth. A 591 (2008) 88. [12] F.H. Attix, Introduction to radiological physics and radiation dosimetry, John Wiley & Sons Inc. (1986). [13] J. H. Lee et al., Development of the multi-purpose gamma-ray detection system consisting of a double-sided silicon strip detector and a 25-segmented germanium detector, Nucl. Phys. A 758 (2005) 150.
–6–
2012 JINST 7 C01009
[2] M. Singh, An electronically collimated gamma camera for single-photon emission computed tomography. Part I: Theoretical considerations and design criteria, Med. Phys. 10 (1983) 421.
