Feasibility Study of an Axially Extendable Multiplex Cylinder PET

IEEE TRANSACTIONS ON NUCLEAR SCIENCE, VOL. 60, NO. 5, OCTOBER 2013 3227

Feasibility Study of an Axially Extendable MultiplexCylinder PET

Eiji Yoshida, Yoshiyuki Hirano, Hideaki Tashima, Naoko Inadama, Fumihiko Nishikido, Hideo Murayama,Hiroshi Ito, and Taiga Yamaya, Member, IEEE

Abstract—Current clinical PET scanners have a 15–22 cmaxial field-of-view (FOV). These scanners image the whole bodyusing six or more bed positions. We designed an axially extend-able multiplex cylinder (AEMC) PET scanner to provide highversatility for clinical and research studies using semiconductorphoto-sensor based, depth-of-interaction (DOI) detectors. Sincesilicon-photomultipliers (Si-PMs) have high gain like conventionalphotomultiplier tubes and a compact design, the Si-PM-baseddetector is particularly expected to enable various new detectorarrangements. The AEMC-PET scanner consists of multiple inde-pendent and laminated detector rings using the DOI detectors. TheAEMC-PET scanner can extend the axial FOV as each stackeddetector ring can be slid sideways. When this PET scanner is usedfor the four-layer DOI detector, its minimum axial FOV is 24 cmand its maximum crystal thickness is 3 cm. On the other hand,the axial FOV can be extended to 96 cm when laminated detectorrings are slid sideways, but the crystal thickness must be 1/4 of3 cm. In this work, we evaluated performance characteristics ofthe PET scanner with a variable axial FOV using Monte Carlosimulation. From the simulation of the 180-cm line source, the96-cm axial FOV was found to have two-fold better sensitivitycompared to the 24-cm axial FOV. For extension of the axial FOV,scatter and attenuation of oblique lines-of-response reduced theyield of true coincidences, but effects of scatter and attenuationwere small. Conclusive results were obtained showing the 52.8-cmaxial FOV yielded an increase in the noise equivalent count rateof approximately 30% relative to the 24-cm axial FOV. We expectthe designed AEMC-PET scanner will provide high versatilityin applications such as for measuring whole-body tracer uptakeswhile keeping the continuous axial FOV; as well, the scan timefor static images will be reduced for a comparable number ofdetectors as conventional PET scanners.

Index Terms—Monte Carlo simulation, PET.

I. INTRODUCTION

C URRENT clinical positron emission tomography (PET)scanners have a 15–22 cm axial field-of-view (FOV).

These scanners image the whole body using six or more bedpositions. Some researchers [1]–[6] have reported on an entirewhole-body PET scanner with an extended axial FOV. Theentire whole-body PET scanner can trace whole-body uptakeimages at the same time and improve sensitivity dramatically.Also, the entire whole-body PET scanner can reduce scan time

Manuscript received February 01, 2013; revised May 31, 2013; acceptedSeptember 01, 2013. Date of publication October 01, 2013; date of currentversion October 09, 2013. This work was supported by the Grant-in-Aid forScientists Research (C) of Kakenhi (24601019).The authors are with the National Institute of Radiological Sciences, Chiba,

Japan (e-mail: [email protected]).Color versions of one or more of the figures in this paper are available online

at http://ieeexplore.ieee.org.Digital Object Identifier 10.1109/TNS.2013.2280633

Fig. 1. Illustration of the AEMC-PET scanner for (a) the dedicated scan modeand (b) the whole-body scan mode.

Fig. 2. Illustrations of the discrete type DOI detector for the AEMC-PETscanner.

in static imaging. Poon et al. [4] have simulated a PET scannerwith about a 2-m axial FOV, which had a noise equivalent countrate (NECR) gain about 30 times higher than that of currentPET scanners. But, the entire whole-body PET scanner is notyet cost effective as a clinical PET scanner. Also, whole-bodyscanning is not needed for every patient.The entire whole-body PET scanner also has many oblique

lines-of-response (LORs). Several characteristics of obliqueLORs, such as their poor spatial resolution and decreased sen-sitivity due to smaller solid angle fraction and increased scatterfraction (SF) and attenuation, should be carefully discussed.Oblique LORs are able to maintain high spatial resolutionusing a depth-of-interaction (DOI) detector [7]–[10]. We havedeveloped a four-layered DOI detector for several dedicatedPET scanners [11], [12]. Another problem is that extremelyoblique LORs may not contribute to imaging performance dueto increased attenuation and scatter in a patient.On the other hand, we have proposed the OpenPET geometry

[13]–[15] which can provide an accessible open space to thepatient during PET scanning. The OpenPET geometry, whichconsists of two or more axially separated detector rings, can ex-tend the axial FOVwith a limited number of detectors. Also, theOpenPET geometry has a variable axial FOV by moving each

0018-9499 © 2013 IEEE

3228 IEEE TRANSACTIONS ON NUCLEAR SCIENCE, VOL. 60, NO. 5, OCTOBER 2013

Fig. 3. Examples of detector ring positions of the AEMC-PET scanner. (a) Minimum axial FOV. (b) Extended axial FOV (2 times). (c) Extended axial FOV (4times). (d) Extended axial FOV (OpenPET).

detector ring independently. However, the OpenPET geometryhas valleys of sensitivity with one appearing on each side ofthe open gap. When the OpenPET geometry has two separatedetector rings, sensitivity profiles are separated into three trian-gular distributions.For an alternative approach, we designed an axially extend-

able multiplex cylinder (AEMC) PET scanner to provide highversatility for clinical and research studies using semiconductorphoto-sensor based, DOI detectors. The AEMC-PET scannerconsists of multiple independent and laminated detector ringsof DOI detectors stacked axially. Each stacked detector ringcan be slid sideways, so the AEMC-PET scanner can extendthe axial FOV with a limited number of detectors the same asthe OpenPET geometry. In this work, we evaluate performancecharacteristics of the PET scanner with a variable axial FOVusing the Geant4 application for tomographic emission (GATE)[16], [17] simulation.

II. MATERIALS AND METHODS

A. Axially Extendable Multiplex Cylinder PET Scanner

Fig. 1 illustrates the AEMC-PET scanner for the dedicatedscan mode and whole-body scan mode. This PET scannerconsists of four independent and laminated detector rings offive DOI detectors stacked axially. Fig. 2 illustrates the discretetype DOI detector design for the AEMC-PET scanner. ThisDOI detector consists of four layers of detectors and eachlayer consists of a 16 16 array of 2.9 2.9 7.5 LSOcrystals and a silicon-photomultiplier (Si-PM) array. SinceSi-PMs have high gain like conventional photomultiplier tubesand a compact design, the Si-PM-based detector is expectedto enable various new detector arrangements. The thinness of

the Si-PM is expected to improve the DOI resolution signif-icantly. Each scintillator array is coupled to the Si-PM array.Depth intervals of each scintillator array are 1.5 cm. TotalLSO thickness of the AEMC-PET scanner is 30 mm. Withthis discrete-type DOI detector, the AEMC-PET scanner canextend the axial FOV as each stacked detector ring can be slidsideways. The AEMC-PET scanner can control the axial FOVfor imaging targets from dedicated studies to entire whole-bodystudies. Fig. 3 shows examples of detector ring positions of theAEMC-PET scanner. This scanner has the minimum axial FOVof 24 cm and maximum crystal thickness of 3 cm (Fig. 3(a)).The axial FOV can be extended to 96 cm by using laminateddetector rings that are slid sideways, but crystal thickness mustbe 1/4 of 3 cm (Fig. 3(c)). Also, the AEMC-PET scanner canbe extended to include several axial FOVs (Fig. 3(b)). In thiscase, the detector ring configuration allows some patterns forthe detector ring positions. In addition, using the OpenPETgeometry, the axial FOV can be extended to 168 cm (Fig. 3(d)).

B. Shift Methods of the Laminated Detector Rings

For the AEMC-PET scanner, sensitivity depends on positionsof the detector rings. We tried two types of shift methods for thelaminated detector rings. Fig. 4 shows the relationship betweenpositions of the laminated detector rings and the total axial FOVfor the shift methods. When the detector ring of the 1st layer isfixed, other detector rings are slid sideways toward the direc-tion of the patient’s legs. The linear shift method is simple, andshift values of the 3rd layer and 4th layer are 2 and 3 times theshift value of the 2nd layer, respectively. When the axial FOVis extended in the linear shift, both ends of the detector ringsbecome 1 layer as shown in Fig. 3(b-1). On the other hand, theAEMC-PET scanner retains more than 2 layers until the 48-cmaxial FOV for the step shift as shown in Fig. 3(b-2). For an axial

YOSHIDA et al.: FEASIBILITY STUDY OF AN AXIALLY EXTENDABLE MULTIPLEX CYLINDER PET 3229

Fig. 4. Relationship between each detector ring position and the total axial FOV for the two types of shift methods. (a) Linear shift. (b) Step shift.

FOV larger than 96 cm, the shift value of the step shift is thesame as that of the linear shift.

C. Simulation Setup

GATE has been used to simulate several PET scanners andreliability and usefulness of the GATE platform have been vali-dated. GATE simulation codes (Version 6.0.0p1) were installedon a 64-bit personal computer (Intel Core i7 3.4 GHz) andwe used them to simulate characteristics of the AEMC-PETscanner. The simulated scanner had a ring diameter of 80 cmas shown in Table I. The scintillator block configuration wasbased on our previous DOI detector [18]. The GATE simulationwas described only for the scintillation block and end-shieldswithout the gantry, Si-PM arrays, and front-end circuits. Aparalysable dead time (250 ns) was applied on the singledata for each DOI detector. Also, DOI detectors were axiallyarranged into a bank with a non-paralysable dead time (256ns) before judging coincidences. The entire whole-body PETscanners have large acceptance angles, which require a widercoincidence timing window to accommodate the time-of-flight(TOF) for long LORs between two detectors. Therefore, thecoincidence time window was 6 ns for detecting oblique LORs.With the same crystal volume, the axial FOV was varied be-tween 24–168 cm and the crystal thickness was varied between7.5 and 30 mm with 7.5 mm increments as shown in Fig. 3.Also, a 1-cm-thick annulus of lead shielding (with an innerdiameter of 70 cm and an outer diameter of 90 cm) was used atthe two axial ends to reduce detection of events from outsidethe FOV.In the discrete-type DOI-PET scanners [19], [20], each layer

of the DOI detector applies an energy window individually.One drawback of our DOI detector is degraded sensitivity dueto inter-crystal scattering, because the AEMC-PET uses thincrystals. When multiple interactions occur between each layer,the energy window tends to reject both events. A softwarecoincidence method can solve this problem, but the number ofdetected data becomes huge. Therefore, it is desirable to treatthe signal of each layer collectively. On the other hand, whenthe axial FOV is extended, some coincidences cause multiple

TABLE IBASIC SPECIFICATIONS OF THE AEMC-PET SCANNER.

coincidences with false LORs. Most multiple coincidencescan be rejected by the energy window. In this simulation,energy signals of each layer were summed as a single detector.When multiple interactions in different layers have not onlyinter-crystal scattering but also multiple coincidences, the DOIdetector adopts the layer with the highest deposited energy as asimple priority rule.

D. Performance Evaluation

A point source and a 180-cm line source of 1 kBq were sim-ulated and the relative sensitivity for the AEMC-PET scannerwas calculated as a function of the axial FOV. These sourceswere set in the center of the FOV. The number of detected


TABLE IISCAN SETUPS FOR MULTIPLE BED POSITIONS. OVERLAP BETWEEN ADJACENT

BED POSITIONS WAS 50%.

true coincidences was computed. In the OpenPET geometry, theaxial gap included virtual slices. Each measurement time was1000 s. Also, sensitivity profiles were calculated for multiplebed positions of several axial FOVs. Total scan time for all sit-uations was 1000 s. Table II shows scan setups for multiple bedpositions. Except for the case of the 96-cm scanner, overlap be-tween adjacent bed positions was 50%.The SF test utilized two types of solid polyethylene cylinder

phantoms (180 cm long and 20 or 30 cm in diameter) with the180-cm line source at 4 cm off-axis. The phantoms were modi-fied from the phantom based on the NEMANU-2 2001 [21] (70cm long and 20 cm in diameter). A phantom was placed in thecenter of the FOV. The SF was calculated as the ratio betweenthe number of scatter coincidences and prompts. To evaluate at-tenuation and scatter, we determined the data-loss fraction ofscatter and attenuation from simulations of a 1-kBq line sourcewith or without the 180-cm cylinder phantoms (20 or 30 cm indiameter). The number of detected true coincidences was com-puted for each phantom. The data-loss fraction of scatter andattenuation was calculated as follows:

(1)where, , and are the true countrates with and without the polyethylene cylinder, respectively.The noise equivalent count rate (NECR) [22] test utilized the

180-cm cylinder phantoms. These phantoms were placed in thecenter of the FOV. The NECR of the AEMC-PET scanner wascalculated as follows:

(2)

where, , and are the true, scatter, and random count rates,respectively. For each axial FOV, peak NECR was estimatedfrom each NECR curve.

Fig. 5. Sensitivities as a function of the axial FOV for the point source.

Fig. 6. Sensitivities as a function of the axial FOV for the 180-cm line source.

III. RESULTS

A. Sensitivity

Fig. 5 shows sensitivities as a function of the axial FOV forthe point source with the two types of shift methods. The max-imum sensitivity was obtained at the minimum axial FOV. Asthe axial FOV was extended, sensitivity rapidly decreased untilthe 96-cm axial FOV was reached. This was because the sensi-tivity depended on both crystal thickness and acceptance angle,and the decrease of detection efficiency for a thin crystal thick-ness exceeded the increase of the acceptance angle for the ex-tended axial FOV. Also, for axial FOVs larger than 96 cm, thecrystal thickness was constant and the sensitivity depended ononly the acceptance angle. And, sensitivity with the linear shiftwas slightly higher than that with the step shift.Fig. 6 shows sensitivities as a function of the axial FOV for

the 180-cm line source with the two types of shift methods. Themaximum sensitivity was obtained at the 96-cm axial FOV. Thisvalue was 2.3 times that for the 24-cm axial FOV. This time, theincrease of occupancy of the line source in the axial FOV ex-ceeded the decrease of detection efficiency due to the thin crystal


Fig. 7. Sensitivity profiles of several axial FOVs for the 180-cm line source.

thickness. Also, for axial FOVs larger than 96 cm, sensitivityhad the same tendency as seen for the point source. Finally, sen-sitivity with the step shift was slightly higher than that with thelinear shift.Fig. 7 shows sensitivity profiles of several axial FOVs for the

180-cm line source. The 24-cm axial FOV had left-right sym-metry, but other configurations had left-right asymmetry due tothe difference in ring diameter of each DOI layer. Also, in the42-, 60-, and 78-cm axial FOVs, the difference in sensitivityprofiles between the linear shift and the step shift methods de-pended on the positions of the detector rings. The AEMC-PETscanner had a continuous axial FOV up to the 96-cm axial FOV.As the axial FOV was further extended, sensitivity profiles withthe axial gap were separated into seven triangular distributions.Fig. 8 shows sensitivity profiles of the multiple bed positions

for the two types of shift methods. Table III shows total sen-sitivity within the 80-cm axial FOV. As the scanner axis FOVwas extended, the total sensitivity increased. The 24-cm scanneraxial FOVs had uniform sensitivity, but the 42-cm, 60-cm and78-cm scanner axial FOVs had non-uniform sensitivity profilesfor left-right asymmetry due to the difference in the ring diam-eter. Sensitivity profiles of the 42-cm scanner axial FOV withthe step shift method, and the 60-cm and 78-cm scanner axialFOVs with the linear shift method were relatively constant com-pared to the respective results for the other shift method.

B. Scatter and Attenuation

Fig. 9 shows SFs as a function of the axial FOV for the180-cm cylinder phantoms using the step shift. When the axialFOV was increased, SF increased slightly. For the 20-cm and30-cm diameter phantoms, SF of the 96-cm axial FOV increasedby 11% and 9.8%, respectively, compared with the SF of the24-cm axial FOV for both phantoms. Fig. 10 shows the data-lossfraction of scatter and attenuation with the 180-cm cylinderphantoms as a function of the axial FOV for the 180-cm linesource using the step shift. When the axial FOV increased, thedata-loss fraction of scatter and attenuation increased slightly.For the 20-cm and 30-cm diameter phantoms, the data-loss frac-tion of scatter and attenuation of the 96-cm axial FOV increasedby 1.2% and 1.6%, respectively, compared with the data-lossfraction of the 24-cm axial FOV for both phantoms.

C. NECR

Fig. 11 shows example NECR curves with several axial FOVsas a function of activity as obtained by the step shift. The 60-cmaxial FOV had the highest NECR curve. On the other hand, the168-cm axial FOV had the lowest NECR curve. Fig. 12 showspeakNECRs as a function of the axial FOV by the step shift. The52.8-cm axial FOV yielded an increase in peak NECR of ap-proximately 30% relative to the 24-cm axial FOV for both phan-


Fig. 8. Sensitivity profiles of the multiple bed positions for (a) the linear shift and (b) the step shift.

TABLE IIITOTAL SENSITIVITY WITHIN THE 80-CM AXIAL FOV.

Fig. 9. SFs as a function of the axial FOV for the 180-cm cylinder phantoms(20 or 30 cm in diameter).

toms. In this case, the activity at the peak NECR was 375 MBq.On the other hand, the 96-cm axial FOV yielded an increase in

Fig. 10. Data-loss fraction of scatter and attenuation with the 180-cm cylinderphantoms (20 or 30 cm in diameter) as a function of the axial FOV for the180-cm line source.

peak NECR of approximately 15% relative to the 24-cm axialFOV for the 20-cm diameter phantom. And, the peak NECR ofthe 96-cm axial FOV was equal to that of the 24-cm axial FOVfor the 30-cm diameter phantom.

IV. DISCUSSION AND CONCLUSION

We designed the AEMC-PET scanner with a limited numberof detectors to extend the axial FOV. The AEMC-PET scanneris expected to provide high versatility in applications, such asdedicated scans and entire whole-body scans, by controlling theaxial FOV for imaging targets. For entire whole-body scans,the AEMC-PET scanner has higher sensitivity than that of thecurrent axial FOV. In other words, the AEMC-PET scanner


Fig. 11. NECR curves with several axial FOVs as a function of activity for the 180-cm cylinder phantoms: (a) 20-cm diameter phantom and (b) 30-cm diameterphantom.

Fig. 12. Peak NECRs as a function of the axial FOV for the 180-cm cylinder phantoms: (a) 20-cm diameter phantom and (b) 30-cm diameter phantom.

can reduce the scan time needed for obtaining static images.To extend the axial FOV, scatter and attenuation of obliqueLORs reduced the yield of true coincidences, but the effects ofscatter and attenuation were small as shown in Figs. 9 and 10.Also, conclusively, NECR performance was improved. Whenthe axial FOV was extended to over 96 cm, the AEMC-PETscanner had the OpenPET geometry. The OpenPET geometrypromises to provide a large axial FOV with the open space, butslightly decreased performance values [23]. On the other hand,we think that the short axial FOV with thick crystals is betterfor dedicated scans such as brain dynamic studies. But, manychallenging issues must be overcome to realize an AEMC-PETscanner for clinical use. For example, the AEMC-PET scannerneeds to move not just a patient bed but a gantry. We will tryto design the gantry for the AEMC-PET scanner and evaluateattenuation of the gantry and the front-end circuit.We mainly evaluated the sensitivity characteristics of sev-

eral positions of detector rings for two types of shift methods.When detector rings were extended, the sensitivity profile hadleft-right asymmetry due to the difference in the ring diameter ofeach DOI layer. Also, sensitivity profiles had slightly differentdistributions between the two types of shift methods. This dif-

ference is attributable to not only detector ring position but alsototal crystal thickness at each axial offset. As a result, the sensi-tivity profile had a complicated distribution as shown in Fig. 7.But, total sensitivities were almost the same for the two typesof shift methods. On one point, however, the step shift methodwas better than the linear shift method for the whole-body scan,because when the axial FOV was extended, both ends of thedetector rings became one layer in the linear shift method, butremained as two layers until the 48-cm axial FOV was reachedin the step shift method.Also, in multiple bed studies for 96-cm scan, the AEMC-PET

had a non-uniform sensitivity profile caused by left-right asym-metry due to the difference in the ring diameter. In this simula-tion, the AEMC-PET scanner was used for all LORs. But, wethink that this non-uniform sensitivity profile could be reducedby optimizing not only positions of the detector rings but alsoby limiting the maximum ring difference. On the other hand, theentire whole-body scan with the single bed position had a trian-gular distribution. In fact, oblique LORs sacrifice the maximumsensitivity in order to obtain a uniform sensitivity profile. Also,oblique LORs cause a parallax error. The DOI detector can re-duce the parallax error, but the AEMC-PET scanner has ex-


tremely oblique LORs. Extremely oblique LORs may not con-tribute to imaging performance. We will try to evaluate imagequality for the AEMC-PET scanner.When the phantom diameter was large and the axial FOV

was extended, the scatter coincidence and attenuation were in-creased. Improvement of the NECR for extension of the axialFOV would be effective if the phantom diameter is small. How-ever, for the 30-cm diameter phantom, the NECR did not dropto that of the 24-cm axial FOV until the axial FOV was 96 cm.To lower costs, the AEMC-PET scanner uses the same num-

bers of LSO scintillators as current PET scanners. However,the discrete type DOI detector needs redundant Si-PMs unlikea conventional PET detector. On the other hand, the discretetype DOI detector has some merits; the thinness of the Si-PMsis expected to improve DOI resolution significantly and thinnerscintillators may have slightly better timing resolution proper-ties. We will develop a prototype of the discrete type DOI de-tector in the near future.In summary, we presented the conceptual design of the

AEMC-PET scanner. The AEMC-PET scanner is expectedto provide high versatility such as for measuring whole-bodytracer uptakes while keeping the continuous axial FOV andfurthermore, the scan time for static images is expected to bereduced for a comparable number of detectors as conventionalPET scanners.

REFERENCES

[1] M. Watanabe, K. Shimizu, T. Omura, N. Sato, M. Takahashi, T. Ko-sugi, K. Ote, A. Katabe, R. Yamada, T. Yamashita, and E. Tanaka, “Ahigh-throughput whole-body PET scanner using flat panel PS-PMTs,”IEEE Trans. Nucl. Sci., vol. 51, no. 3, pp. 796–800, 2004.

[2] L. Eriksson, M. Conti, C. L. Melcher, D. W. Townsend, M. Eriksson,H. Rothfuss,M. E. Casey, and B. Bendriem, “Towards sub-minute PETexamination times,” IEEE Trans. Nucl. Sci., vol. 58, no. 1, pp. 76–81,2011.

[3] P. Crespo, J. Reis, M. Couceiro, A. Blanco, N. C. Ferreira, R. F. Mar-ques, P. Martins, and P. Fonte, “Whole-body single-bed time-of-flightRPC-PET: Simulation of axial and planar sensitivities with NEMA andanthropomorphic phantoms,” IEEE Trans. Nucl. Sci., vol. 59, no. 3, pp.520–529, 2012.

[4] J. K. Poon, M. L. Dahlbom, W. W. Moses, K. Balakrishnan, W. Wang,S. R. Cherry, and R. D. Badawi, “Optimal whole-body PET scannerconfigurations for different volumes of LSO scintillator: A simulationstudy,” Phys. Med. Biol., vol. 57, no. 13, pp. 4077–4094, Jun. 2012.

[5] W. H. Wong, J. Uribe, H. Li, H. Baghaei, Y. Wang, M. Aykac, Y. Liu,T. Xing, D. Bilgen, and R. Farrell, “The design of a high-resolutiontransformable wholebody PET camera,” IEEE Trans. Nucl. Sci., vol.49, no. 5, pp. 2079–2084, Oct. 2002.

[6] L. R. MacDonald, R. L. Harrison, A. M. Alessio, W. C. J. Hunter, T. K.Lewellen, and P. E. Kinahan, “Effective count rates for PET scannerswith reduced and extended axial field of view,” Phys. Med. Biol., vol.56, no. 12, pp. 3629–3643, May 2011.

[7] M. L. Dahlbom, L. R. MacDonald, M. Schmand, L. Eriksson, M. An-dreaco, and C. W. Williams, “A YSO/LSO phoswich array detector forsingle and coincidence photon imaging,” IEEE Trans. Nucl. Sci., vol.45, no. 3, pp. 1128–1132, 1998.

[8] M. Schmand, L. Eriksson, M. E. Casey, M. S. Andreaco, C. Melcher,K. Wienhard, G. Flugge, and R. Nutt, “Performance results of a newDOI detector block for a high resolution PET-LSO research tomographHRRT,” IEEE Trans. Nucl. Sci., vol. 45, no. 6, pp. 3000–3006, 1998.

[9] J. Seidel, J. J. Vaquero, S. Siegel, W. R. Gandler, and M. V. Green,“Depth identification accuracy of a three layer phoswich PET detectormodule,” IEEE Trans. Nucl. Sci., vol. 46, no. 3, pp. 485–490, 1999.

[10] T. Tsuda, H. Murayama, K. Kitamura, N. Inadama, T. Yamaya, E.Yoshida, F. Nishikido, M. Hamamoto, H. Kawai, and Y. Ono, “Perfor-mance evaluation of a subset of a four-layer LSO detector for a smallanimal DOI PET scanner: JPET-RD,” IEEE Trans. Nucl. Sci., vol. 53,no. 1, pp. 35–39, 2006.

[11] E. Yoshida, K. Kitamura, T. Tsuda, K. Shibuya, T. Yamaya, N. In-adama, T. Hasegawa, and H. Murayama, “Energy spectra analysis ofthe four-layer DOI detector for the brain PET scanner: JPET-D4,”Nucl.Inst. Methods Phys. Res., A, vol. 557, no. 2, pp. 664–669, Feb. 2006.

[12] T. Yamaya, E. Yoshida, T. Inaniwa, S. Sato, Y. Nakajima, H. Wak-izaka, D. Kokuryo, A. Tsuji, T. Mitsuhashi, H. Kawai, H. Tashima, F.Nishikido, N. Inadama, H. Murayama, H. Haneishi, M. Suga, and S.Kinouchi, “Development of a small prototype for a proof-of-conceptof OpenPET imaging,” Phys. Med. Biol., vol. 56, no. 4, pp. 1123–1137,Jan. 2011.

[13] T. Yamaya, T. Inaniwa, S. Minohara, E. Yoshida, N. Inadama, F.Nishikido, K. Shibuya, C. F. lam, and H. Murayama, “A proposal ofan open PET geometry,” Phys. Med. Biol., vol. 53, no. 3, pp. 757–773,Jan. 2008.

[14] T. Yamaya, T. Inaniwa, E. Yoshida, F. Nishikido, K. Shibuya, N. In-adama, and H. Murayama, “Simulation studies of a new ‘OpenPET’geometry based on a quad unit of detector rings,” Phys. Med. Biol.,vol. 54, no. 5, pp. 1223–1233, Jan. 2009.

[15] T. Yamaya, E. Yoshida, N. Inadama, F. Nishikido, K. Shibuya, M.Higuchi, and H. Murayama, “A Multiplex ‘OpenPET’ geometry toextend axial FOV without increasing the number of detectors,” IEEETrans. Nucl. Sci., vol. 56, no. 5, pp. 2644–2650, 2009.

[16] S. Jan, G. Santin, D. Strul, S. Staelens, K. Assié, D. Autret, S. Avner,R. Barbier, M. Bardiès, P. M. Bloomfield, D. Brasse, V. Breton, P.Bruyndonckx, I. Buvat, A. F. Chatziioannou, Y. Choi, Y. H. Chung, C.Comtat, D. Donnarieix, L. Ferrer, S. J. Glick, C. J. Groiselle, D. Guez,P. F. Honore, S. Kerhoas-Cavata, A. S. Kirov, V. Kohli, M. Koole, M.Krieguer, D. J. J. van der Laan, F. Lamare, G. Largeron, C. Lartizien,D. Lazaro, M. C. Maas, L. Maigne, F. Mayet, F. Melot, C. Merheb, E.Pennacchio, J. Perez, U. Pietrzyk, F. R. Rannou, M. Rey, D. R. Schaart,C. R. Schmidtlein, L. Simon, T. Y. Song, J.-M. Vieira, D. Visvikis, R.V. de Walle, E. Wieërs, and C. Morel, “GATE: A simulation toolkit forPET and SPECT,” Phys. Med. Biol., vol. 49, no. 19, pp. 4543–4561,Sep. 2004.

[17] S. Jan, D. Benoit, E. Becheva, T. Carlier, F. Cassol, P. Descourt, T.Frisson, L. Grevillot, L. Guigues, L. Maigne, C. Morel, Y. Perrot, N.Rehfeld, D. Sarrut, D. R. Schaart, S. Stute, U. Pietrzyk, D. Visvikis, N.Zahra, and I. Buvat, “GATE V6: A major enhancement of the GATEsimulation platform enablingmodelling of CT and radiotherapy,” Phys.Med. Biol., vol. 56, no. 4, pp. 881–901, Jan. 2011.

[18] E. Yoshida, S. Kinouchi, H. Tashima, F. Nishikido, N. Inadama, H.Murayama, and T. Yamaya, “System design of a small OpenPET pro-totype with 4-layer DOI detectors,” Radiol. Phys. Technol., vol. 5, no.1, pp. 92–97, Jan. 2012.

[19] D. P.McElroy,W. Pimpl, B. J. Pichler, M. Rafecas, T. Schüler, and S. I.Ziegler, “Characterization and readout of MADPET-II detector mod-ules: Validation of a unique design concept for high resolution smallanimal PET,” IEEE Trans. Nucl. Sci., vol. 52, no. 1, pp. 199–204, 2005.

[20] T. Shiga, Y. Morimoto, N. Kubo, N. Katoh, C. Katoh, W. Takeuchi, R.Usui, K. Hirata, S. Kojima, K. Umegaki, H. Shirato, and N. Tamaki, “Anew PET scanner with semiconductor detectors enables better identi-fication of intratumoral inhomogeneity,” J. Nucl. Med., vol. 50, no. 1,pp. 148–155, Dec. 2008.

[21] Performance Measurements of Positron Emission TomographsNEMA, Nat. Elect. Manufact. Assoc., Rosslyn, VA, 2001, NEMAStandards Pub. NU 2-2001.

[22] S. C. Strother, M. E. Casey, and E. J. Hoffman, “Measuring PETscanner sensitivity: Relating countrates to image signal-to-noise ratiosusing noise equivalents counts,” IEEE Trans. Nucl. Sci., vol. 37, no.2, pp. 783–788, 1990.

[23] E. Yoshida, T. Yamaya, K. Shibuya, F. Nishikido, N. Inadama, and H.Murayama, “Simulation study on sensitivity and count rate character-istics of ‘OpenPET’ geometries,” IEEE Trans. Nucl. Sci., vol. 57, no.1, pp. 111–116, 2010.

Feasibility Study of an Axially Extendable Multiplex Cylinder PET

Documents

Transcript of Feasibility Study of an Axially Extendable Multiplex Cylinder PET