A GATE Simulation Study of the Siemens Biograph DUO PET/CT System

Dimitrios Nikolopoulos, Sofia Kottou, Nikolaos Chatzisavvas, Xenophon Argyriou, Emannouel Vlamakis, Panayiotis Yannakopoulos, Anna Louizi

Department of Engineering of Electronic and Computer Systems, Technological Educational Institute (TEI) of Piraeus, Athens, Greece.
Department of Physics, Chemistry and Material Science, Technological Educational Institute (TEI) of Piraeus, Athens, Greece.
Medical Physics Department, Medical School, University of Athens, Athens, Greece.

DOI: 10.4236/ojrad.2013.32009                

Abstract

This is a GATE-simulation study of the Siemens Biograph DUO PET/CT system. It reports effects of changes in the thickness of the employed Lutetium Oxyorthosilicate(LSO) detectors. The PET/CT, a human body phantom and a cylindrical F-18 FDG source were simulated. Validation measurements were conducted. The results indicate that LSO thickness increase degrades spatial resolution, improves relative energy resolution from 9.0% to 11.3% and increases signal-to-noise-ratio from 0.81 to 1.17. Thicker LSO crystals present greater axial sensitivity so as the detection efficiency of PET would be significantly enhanced.

Keywords

Monte-Carlo; GATE; PET; Siemens Biograph DUO PET/CT

1. Introduction

Positron emission tomography (PET) is a very powerful medical diagnostic method to observe the metabolism, blood flow, neurotransmission and handling of important biochemical entities [1]. Among the various scintillation detectors employed in commercial PET systems, Be₄Ge₃O₁₂ (BGO) was for a long time the state of the art [2,3]. LuSiO₅ (LSO) has become the best competitor of BGO [1,3-5], mainly due to its high detection efficiency. Other PET systems employ other scintillators, such as Gd₂SiO₅ (GSO), LuAlO₃ (Lu AP), YAlO₃ (YAP) and Y₃Al₅O₁₂ (YAG) [1,3,4]. What’s remarkable is the recent interest in introducing new scintillators and detector designs for PET [2,3,6-10]. Significant improvements have been achieved in the overall PET imaging technology [1,3,4], e.g. algorithms for statistical effects, scatter and random coincidences, faster detector electronics and better reconstruction algorithms [1,3,11,12]. Modern PET scanners incorporate computed tomography (CT) systems to achieve more accurate anatomical localisation of functional abnormalities [12]. The hybrid PET/CT systems eliminate lengthy PET transmission scans and generate complex three dimensional images within few minutes. This improves count-rate, spatial resolution and signal-to-noise ratio (SNR) [2,3,13]. At the same time it enhances clinical conditions, diagnosis, follow-up and therapy [12]. PET/CT technology is undergoing a rapid evolution. As the current technology becomes more widespread, it is likely that there will be a demand for PET designs of better performance and less cost [1,3]. This intensifies the interest for investigations on already employed PET scintillators [5,7,11,14-16] and in seeking applicability of new detector concepts. In designing and evaluating scintillation detectors for PET, it is of significance to determine the various phenomena that affect radiation detection [3,17]. What’s important is the emission and re-absorption of scatter and characteristic X-ray fluorescence radiation, bremsstrahlung and Auger and Koster-Kronig electrons [18]. This is because these phenomena occur apart from the primary interaction point and, as a result, render degradation of spatial resolution and image contrast [19]. In simulating the stochastic processes involved in radiation detection, the Monte Carlo techniques constitute a very efficient tool [4,17]. Several general Monte Carlo packages are available (e.g. PENELOPE, MCNP, EGSnrc MP, GEANT4) [17,19]. Their design is for complex and general geometries of particle showers; however, non-trivial coding is needed. Especially for PET, GATE (GEANT4 Application for Tomographic Emission) is more frequently used because of its flexibility for Tomographic simulations [11].

2. Materials and Methods

The present study focused on the Siemens Biograph DUO PET/CT. The study extended previous validated work [11,14,20] and performed a simulation of the entire PET/CT scanner using GATE. For the purpose of the study, a human body phantom and a cylindrical F-18 FDG source were also simulated. The work emphasises on changes that will be potentially induced if the thickness of the LSO detectors of the PET scanner is altered. For further validation, new experimental measurements were taken. The experimental setup was modeled with GATE and the corresponding outputs were compared with the actual measurements.

2.1. Description of the Simulated Scanner

The simulated Biograph DUO PET/CT is installed in the Diagnostic and Therapeutic Center of Athens, “Hygeia” (Greece). The scanner comprises a dual-slice Siemens Emotion CT scanner in tandem with an ECAT HR⁺ PET scanner. The HR⁺ has no septa and operates entirely in 3D mode. A new patient bed design allows a combined scan range for both PET and CT.

PET and CT acquisition and reconstruction run under a single protocol on one workstation [21]. The CT images are used for the correction of the PET data due to attenuation and scatter. The corrected PET data are reconstructed with the Fourier re-binning (FORE) algorithm and the attenuation-weighted ordered subset EM (AWOSEM) algorithm. The complete acquisition of both PET and CT data takes less than 30 min and the fused images are available for viewing within 5 min after the completion of the scan. The images are viewed on a separate fused image display station [20,21].

The detectors of the Siemens Biograph PET scanner are organised in: 1) buckets, 2) heads, 3) blocks and 4) arrays. The buckets are composed of sets of four heads. Each head contains three blocks and each block contains an 8 × 8 array of LSO crystals. The detector blocks are coupled to sets of four photomultiplier (PMT) tubes. The entire block-photomultiplier arrangement is repeated three times in stacked detector rings. The whole detector settlement finally sums up 48 heads and 144 blocks of 9216 LSO crystals coupled to 576 photomultiplier tubes.

The PET and CT scanners are separated through 2.5 cm thick lead arcs arranged around the detector setup [20,21]. Figure 1(a) presents the simulated scanner. Figures 1(b)-(k) illustrate schematically the parts of the PET detectors. Figure 1(l) presents the actual crystal-PMT

Figure 1. The investigated PET/CT Biograph DUO LSO scanner. (a) The actual installation in the Diagnostic and Therapeutic Center of Athens, “Hygeia” (Greece); (b)-(k) Geometrical simulation of various parts of the scanner; (l) The real 8 × 8 LSO array of the PET system; (m) The actual detector assembly; (n) One real detection block of the PET scanner system.

assembly and Figure 1(m) the real LSO detector assembly for coincidence measurements.

The dimensions of the actual PET/CT components are the following: 1) Entry port diameter of 70 cm for both PET and CT; 2) Overall tunnel length of 110 cm; 3) Combined scan range of 145 cm; 4) PET gantry radii between 35.0 cm and 53.5 cm and gantry height of 18.8 cm; 5) Buckets dimensions of 12 × 21.6 × 16.18 cm³; 6) Heads dimensions of 2 × 5.393 × 16.180 cm³; 7) Block dimensions of 5.393 × 5.393 × 10.5 cm³; 8) Arrays consisting of LSO crystals of dimensions 0.645 × 0.645 × 2.5 cm³; and 9) Photomultiplier tubes of radius 1.27 cm and height 7.9 cm.

2.2. Description of GATE

GATE is a GEANT4 based Monte-Carlo platform adapted to the field of Nuclear Medicine. Through a dedicated script language, it may simulate the passage of particles through matter and electromagnetic fields providing different levels of description, analysis and visualisation. It may simulate detector and source kinetics and other time-dependent phenomena rendering hence, the coherent description of acquisition processes and detector output pulses.

Detector response is modeled by a chain of processing modules comprising 1) the Adder which regroups the hits per volume into a pulse, 2) the Readout which regroups the pulses per block, 3) the Energy Response which simulates a Gaussian blurring of the energy spectrum of a pulse after the readout module, 4) the Spatial Response which provides the coincidence events and the lines of response (LOR) needed for the image reconstruction, 5) the Threshold Electronics which provide the cut-off energy windows and 6) the Dead Time which defines the dead-time behavior of the counting system.

2.3. Description of the Simulation

The GATE codes simulated the following parts: 1) the entire PET detector arrangement; 2) the light guides, photomultiplier tubes and related electronics; 3) the coincidence circuits and processors; 4) the digitizer; 5) the time-delay of PET; 6) the data processing systems; 7) the examination bed; 8) the PET gantry; 9) the PET motions (gantry, bed); 10) the shielding between PET and CT; 11) the shielding of the room; and 12) the CT image reconstruction process. Noteworthy is that CT was simulated so as to reproduce in the most efficient way the image acquisition and processing techniques followed during PET scanning.

Additionally software phantoms were simulated consisted of 1) a cylindrical source phantom of radius 1 mm and height 15 cm, homogeneously filled with F-18 FDG and 2) a human body phantom.

This phantom consisted of an ellipsoid of 8 cm minimum and 15 cm maximum radius mimicking the human main body (b-1), two cylinders of radius 5 cm and height 30 cm mimicking the human hands (b-2), a sphere of 14 cm radius mimicking the human head (b-3) and a cylindrical F-18 FDG source of 0.5 cm radius and 5 cm height settled around the centre of the ellipsoid-human torso. All software phantoms were computationally arranged in a manner that the central axis of the enclosed F-18 FDG source was aligned to the central axis of the gantry.

During simulation all interaction phenomena were allowed to occur with the following parameters: 1) crystal energy blurring: resolution of 0.26 at 511 keV; 2) detector characteristics: gamma ray absorption linear coefficient of 0.98, light output of 30,000 photons/MeV, intrinsic resolution of 0.088 and transfer efficiency coefficient of 0.28; 3) energy window: between 250 and 650 keV; 4) time resolution: coincidence window, dead time window of 120 ns and dead time offset of 700 ns; 5) F-18 FDG source half life of 6586.2 s; and 6) slice time of 1 s and acquisition time of 10 s.

Parameters 1) and 2) are equal to the manufacturer values. Parameters 3) to 4) are the values adopted during operation.

2.4. Validation

Validation measurements were derived with a cylindrical F-18 FDG source of 1 mm radius, 15 cm height and 29.6 MB qactivity for acquisition time of 10 s. The source was placed at the centre of the PET gantry. The validation measurements were imitated with the modeled source (software phantom a) which was computationally settled at the gantry’s centre. In order to increase computational accuracy, the modeled activity was set to 100 MBq. For further mimicking, simulated acquisition time was set to 10 s and the profiles of the four photomultiplier tubes were also generated through modeling for comparison. Further validation was performed by comparing simulation results of software phantoms (a) and (b) with those anticipated from the physics of PET imaging. Figures 2(a)-(c) present the actual experimentation during validation measurements. Figure 2(d) shows the simulated source (software phantom a). Figures 2(e) and (f) present two views of the human phantom (software phantom b).

3. Results and Discussion

Figure 3 presents the results of the additional validation experiments together with results of the similar software phantom of Figure 2(d). The normalised energy spectra of the four photomultiplier tubes in block 0 and bucket 0 (Figure 3(Ia)) were the actual spectra provided by the Siemens Biograph DUO PET/CT from validation measurements.

Conflicts of Interest

The authors declare no conflicts of interest.

References

[1]	C. W. E. Van Eijk, “Inorganic Scintillators in Medical Imaging,” Physics in Medicine and Biology, Vol. 47, No. 8, 2002, pp. R85-R106. doi:10.1088/0031-9155/47/8/201
[2]	T. K. Lewellen, “Recent Developments in PET Detector Technology,” Physics in Medicine and Biology, Vol. 53, No. 17, 2008, pp. R287-R317. doi:10.1088/0031-9155/53/17/R01
[3]	C. W .E. Van Eijk, “Radiation Detector Developments in Medical Applications: Inorganic Scintillators in Positron Emission Tomography,” Radiation Protection Dosimetry, Vol. 129, No. 1-3, 2008, pp. 13-21. doi:10.1093/rpd/ncn043
[4]	M. Nikl, “Scintillation Detectors for X-Rays,” Measurement Science and Technology, Vol. 17, No. 4, 2006, pp. R37-R54. doi:10.1088/0957-0233/17/4/R01
[5]	J. K. Poon, M. L. Dahlbom, W. W. Moses, K. Balakrishnan, W. Wang, S. R. Cherry and R. D. Badawi, “Optimal Whole-Body PET Scanner Configurations for Different Volumes of LSO Scintillator: A Simulation Study,” Physics in Medicine and Biology, Vol. 57, No. 13, 2012, pp. 4077-4094. doi:10.1088/0031-9155/57/13/4077
[6]	P. Geramifar, M. R. Ay, M. Shamsaie Zafarghandi, S. Sarkar, G. Loudos and A. Rahmim, “Investigation of Time of-Flight Bene?ts in an LYSO-Based PET/CT Scanner: A Monte Carlo Study Using GATE,” Nuclear Instruments and Methods in Physics Research Section A, Vol. 641, No. 1, 2011, pp.121-127. doi:10.1016/j.nima.2011.03.030
[7]	A. R. Karimian and C. J. Thompson, “Assessment of a New Scintillation Crystal (LaBr3) in PET Scanners Using Monte Carlo Method,” Nukleonika, Vol. 53, No 1, 2008, pp. 3-6.
[8]	T. Nakamori, T. Kato, J. Kataoka, T. Miura, H. Matsuda, K. Sato, Y. Ishikawa, K. Yamamura, N. Kawabata, H. Ikeda, G. Satoc and K. Kamadad, “Development of a Gamma-Ray Imager Using a Large Area Monolithic 4×4 MPPC Array for a Future PET Scanner,” Journal of Instrumentation, Vol. 7, No. 1, 2012, pp. 1-13.
[9]	O. Mineev, Y. Kudenko, Y. Musienko, I. Polyansky and N. Yershov, “Scintillator Detectors with Long WLS Fibers and Mulit-Pixel Photodiodes,” Journal of Instrumentation, Vol. 6, No. 12, 2011, pp. 1-9.
[10]	A. Vandenbroucke, A. M. K. Foudray, P. D. Olcott and C. S. Levin, “Performance Characterization of a New High Resolution PET Scintillation Detector,” Physics in Medicine and Biology, Vol. 56, No. 3, 2011, pp. 4135-4145.
[11]	P. Gonias, N. Bertsekas, N. A. Karakatsanis, G. Saatsakis, D. Nikolopoulos, X. Tsantilas, G. Loudos, N. Sakellios, N. Gaitanis, A. Papaspyrou, L. Daskalakis, A. Liaparinos, D. Cavouras, I. Kandarakis and G. S. Panayiotakis, “Validation of a GATE Model for the Simulation of the Siemens PET/CT Biograph 6 Scanner,” Nuclear Instruments and Methods in Physics Research Section A, Vol. 571, No. 1-2, 2007, pp. 263-266. doi:10.1016/j.nima.2006.10.078
[12]	D. W. Townsend, “Physical Principles and Technology of Clinical PET Imaging,” ANNALS Academy of Medicine, Vol. 33, No. 2, 2004, pp. 133-145.
[13]	D. L. Bailey and S. R. Meikle, “A Convolution-Subtraction Scatter Correction Method for 3D PET,” Physics in Medicine and Biology, Vol. 39, No. 3, 1994, pp. 411-424. doi:10.1088/0031-9155/39/3/009
[14]	N. Karakatsanis, N. Sakellios, N. X. Tsantilas, N. Dikaios, C. Tsoumpas, D. Lazaro, G. Loudos, C. R. Schmidtlein, A. Louizi, J. Valais, D. Nikolopoulos, J. Malamitsi, J. Kandarakis and K. Nikita, “Comparative Evaluation of Two Commercial PET Scanners, ECAT EXACT HR+ and Biograph 2, Using GATE,” Nuclear Instruments and Methods in Physics Research Section A, Vol. 571, No. 2, 2006, pp. 368-372. doi:10.1016/j.nima.2006.08.110
[15]	S. Shimizu, C. M. Pepin and R. Lecomte, “Assessment of Lu1:8Gd0:2SiO5 (LGSO) Scintillators with APD Readout for PET/SPECT/CT Detectors,” IEEE Transactions on Nuclear Science, Vol. 57, No. 3, 2010, pp. 1512-1517. doi:10.1109/TNS.2010.2048435
[16]	N. Zeraatkar, M. R. Ay, A. R. Kamali-Asl and H. Zaidi, “Accurate Monte Carlo modeling and Performance Assessment of the X-PET? Subsystem of the FLEX Triumph? Preclinical PET/CT Scanner,” Medical Physics, Vol. 38, No. 3, 2011, pp. 1217-1225. doi:10.1118/1.3547721
[17]	D. Nikolopoulos, I. Kandarakis, X. Tsantilas, I. Valais, D. Cavouras and A. Louizi, “Comparative Study of the Radiation Detection Efficiency of LSO, LuAP, GSO and YAP Scintillators for Use in Positron Emission Imaging (PET) via Monte-Carlo Methods,” Nuclear Instruments and Methods in Physics Research Section A, Vol. 569, No. 2, 2006, pp. 350-354. doi:10.1016/j.nima.2006.08.033
[18]	D. Nikolopoulos, I. Kandarakis, D. Cavouras, A. Louizi and C. Nomicos, “Investigation of Radiation Absorption and X-Ray Fluorescence of Medical Imaging Scintillators by Monte Carlo Methods,” Nuclear Instruments and Methods in Physics Research Section A, Vol. 565, No. 2, 2006, pp. 821-832. doi:10.1016/j.nima.2006.05.170
[19]	J. M. Boone, J. A. Seibert, J. M. Sabol and M. Tecotzky, “A Monte Carlo Study of X-Ray Fluorescence in X-Ray Detectors,” Journal of Medical Physics, Vol. 26, No. 6, 1999, pp. 905-916. doi:10.1118/1.598612
[20]	X. Tsantilas, A. Louizi, I. Valais, D. Nikolopoulos, N. Sakellios, N. Karakatsanis, G. Loudos, K. Nikita, J. Malamitsi and I. Kandarakis, “Simulation of Commercial PET Scanners with GATE Monte-Carlo Simulation Package,” Journal of Biomedicine and Biotechnology, Vol. 50, Suppl. 1, 2005, pp. 114-115.
[21]	D. W. Townsend and T. Beyer, “A Combined PET/CT Scanner: The Path to True Image Fusion,” British Journal of Radiology, Vol. 75, Suppl. 9, 2002, pp. S24-S30.
[22]	G. F. Knoll, “Radiation Detection and Measurement,” John Wiley & Sons, New York, 1979.
[23]	C. Lartizien, C. Kuntne, A. L. Goertzen, A. C. Evans and A. Reilhac, “Validation of PET-SORTEO Monte Carlo Simulations for the Geometries of the MicroPET R4 and Focus 220 PET Scanners,” Physics in Medicine and Biology, Vol. 52, No. 16, 2007, pp. 4845-4862. doi:10.1088/0031-9155/52/16/009
[24]	C. Merheb, Y. Petegnief and J. N. Talbot, “Full Modelling of the MOSAIC Animal PET System Based on the GATE Monte Carlo Simulation Code,” Physics in Medicine and Biology, Vol. 52, No. 3, 2007, pp. 563-576. doi:10.1088/0031-9155/52/3/002
[25]	D. A. B. Bonifacio, N. Belcari, S. Moehrs, M. Moralles, V. Rosso, S. Vecchio and A. Del Guerra, “A Time Ef?cient Optical Model for GATE Simulation of a LYSO Scintillation Matrix Used in PET Applications,” IEEE Transactions on Nuclear Science, Vol. 57, No. 5, 2010, pp. 2483-2489. doi:10.1109/TNS.2010.2062536
[26]	D. J. Vander Laan, D. R. Schaart, M. C. Maas, F. J. Beekman, P. Bruyndonckx and C. W. E. Van Eijk, “Optical Simulation of Monolithic Scintillator Detectors Using GATE/GEANT4,” Physics in Medicine and Biology, Vol. 55, No. 6, 2010, pp. 1659-1675. doi:10.1088/0031-9155/55/6/009
[27]	C. R. Schmidtlein, “Validation of GATE Monte Carlo Simulations of the GE Advance/Discovery LS PET Scanners,” Medical Physics, Vol. 33, No. 1, 2006, pp. 198-208. doi:10.1118/1.2089447
[28]	C. Michel, L. Eriksson, H. Rothfuss and B. Bendriem, “In?uence of Crystal Material on the Performance of the HiRez 3D PET Scanner: A Monte Carlo Study,” IEEE Nuclear Science Symposium Conference, San Diego, 2006, pp. 2528-2531.
[29]	S. Jan, C. Comtat, D. Strul, G. Santin and R. Trbossen, “Monte Carlo Simulation for the ECAT EXACT HR+ System Using GATE,” IEEE Transactions on Nuclear Science, Vol. 52, No. 3, 2005, pp. 627-633. doi:10.1109/TNS.2005.851461
[30]	T. K. Lewellen, “The Challenge of Detector Designs for PET,” American Journal of Roentgenology, Vol. 195, No. 2, 2010, pp. 301-309. doi:10.2214/AJR.10.4741
[31]	G. L. Brownell, J. A. Correia and R. G. Zamenhof, “Positron Instrumentation,” In: J. H. Lawrence and T. F. Budinger, Eds., Recent Advances in Nuclear Medicine, Grune & Stratton, New York, 1978, pp. 1-49.
[32]	S. C. Strother, M. E. Casey and E. J. Hoffman, “Measuring PET Scanner Sensitivity: Relating Countrates to Image Signal-to-Noise Ratios Using Noise Equivalent Counts,” IEEE Transactions on Nuclear Science, Vol. 37, No. 2, 1990, pp. 783-788. doi:10.1109/23.106715
[33]	J. S. Karp, S. Surti, M. E. Daube-Witherspoon and G. Muehllehner, “Bene?t of Time-of-Flight in PET: Experimental and Clinical Results,” Journal of Nuclear Medicine, Vol. 49, No. 3, 2008, pp. 462-470. doi:10.2967/jnumed.107.044834
[34]	G. El Fakhri, S. Surti, C. M. Trott, J. Scheuermann and J. S. Karp, “Improvement in Lesion Detection with Whole Body Oncologic Time-of-Flight PET,” Journal of Nuclear Medicine, Vol. 52, No. 3, 2011, pp. 347-353. doi:10.2967/jnumed.110.080382
[35]	S. R Cherry, “In Vivo Molecular and Genomic Imaging: New Challenges for Imaging Physics,” Physics in Medicine and Biology, Vol. 49, No. 3, 2004, pp. R13-R48. doi:10.1088/0031-9155/49/3/R01
[36]	L. Pan, J. He and L. Ma, “An Initial Simulation Study of PET Imaging by GATE,” 2010 2nd International Conference on Information Science and Engineering (ICISE), Hangzhou, 2010.
[37]	S. Tavernier, P. Bruyndonckx, S. Leonard and O. Devroede, “A High-Resolution PET Detector Based on Continuous Scintillators,” Nuclear Instruments and Methods in Physics Research Section A, Vol. 537, 2005, pp. 321-325. doi:10.1016/j.nima.2004.08.035
[38]	C. Yong, J. Jin Ho, C. Yong Hyun, O. Devroede, M. Krieguer, P. Bruyndonckx and S. Tavernier, “Optimization of LSO/LuYAP phoswich Detectzor for Small Animal PET,” Nuclear Instruments and Methods in Physics Research Section A, Vol. 571, No. 3, 2007, pp. 669-675. doi:10.1016/j.nima.2006.10.293