-
A small animal PET based on GAPDs and charge signal transmission
approach for hybrid
PET-MR imaging
This article has been downloaded from IOPscience. Please scroll
down to see the full text article.
2011 JINST 6 P08012
(http://iopscience.iop.org/1748-0221/6/08/P08012)
Download details:
IP Address: 171.65.81.183
The article was downloaded on 22/09/2011 at 21:52
Please note that terms and conditions apply.
View the table of contents for this issue, or go to the journal
homepage for more
Home Search Collections Journals About Contact us My
IOPscience
http://iopscience.iop.org/page/termshttp://iopscience.iop.org/1748-0221/6/08http://iopscience.iop.org/1748-0221http://iopscience.iop.org/http://iopscience.iop.org/searchhttp://iopscience.iop.org/collectionshttp://iopscience.iop.org/journalshttp://iopscience.iop.org/page/aboutioppublishinghttp://iopscience.iop.org/contacthttp://iopscience.iop.org/myiopscience
-
2011 JINST 6 P08012
PUBLISHED BY IOP PUBLISHING FOR SISSA
RECEIVED: June 29, 2011ACCEPTED: August 9, 2011
PUBLISHED: August 24, 2011
A small animal PET based on GAPDs and chargesignal transmission
approach for hybrid PET-MRimaging
Jihoon Kang,a,b Yong Choi,a,1 Key Jo Hong,a Wei Hu,a,b Jin Ho
Jung,a
Yoonsuk Huha,b and Byung-Tae KimbaDepartment of Electronic
Engineering, Sogang University,1 Shinsu-Dong, Mapo-Gu, Seoul
121-742, Republic of Korea
bDepartment of Nuclear Medicine, Samsung Medical
Center,Sungkyunkwan University School of Medicine,50 Ilwon-Dong,
Gangnam-Gu, Seoul 135-710, Republic of Korea
E-mail: [email protected]
ABSTRACT: Positron emission tomography (PET) employing
Geiger-mode avalanche photodiodes(GAPDs) and charge signal
transmission approach was developed for small animal imaging.
Ani-mal PET contained 16 LYSO and GAPD detector modules that were
arranged in a 70 mm diameterring with an axial field of view of 13
mm. The GAPDs charge output signals were transmittedto a
preamplifier located remotely using 300 cm flexible flat cables.
The position decoder circuits(PDCs) were used to multiplex the PET
signals from 256 to 4 channels. The outputs of the PDCswere
digitized and further-processed in the data acquisition unit. The
cross-compatibilities of thePET detectors and MRI were assessed
outside and inside the MRI. Experimental studies of thedeveloped
full ring PET were performed to examine the spatial resolution and
sensitivity. Phantomand mouse images were acquired to examine the
imaging performance. The mean energy and timeresolution of the PET
detector were 17.6% and 1.5 ns, respectively. No obvious
degradation onPET and MRI was observed during simultaneous PET-MRI
data acquisition. The measured spatialresolution and sensitivity at
the CFOV were 2.8 mm and 0.7%, respectively. In addition, a 3
mmdiameter line source was clearly resolved in the hot-sphere
phantom images. The reconstructedtransaxial PET images of the mouse
brain and tumor displaying the glucose metabolism patternswere
imaged well. These results demonstrate GAPD and the charge signal
transmission approachcan allow the development of high performance
small animal PET with improved MR compatibil-ity.
KEYWORDS: Gamma camera, SPECT, PET PET/CT, coronary CT
angiography (CTA); Multi-modality systems
1Corresponding author.
c 2011 IOP Publishing Ltd and SISSA
doi:10.1088/1748-0221/6/08/P08012
mailto:[email protected]://dx.doi.org/10.1088/1748-0221/6/08/P08012
-
2011 JINST 6 P08012
Contents
1 Introduction 1
2 Materials and methods 22.1 System description 22.2 Performance
measurement of the PET detector modules outside MRI 32.3
Characterization of the cross-compatibility of PET detector module
and MRI 42.4 Phantom and small animal imaging of the full ring PET
4
2.4.1 Spatial resolution and sensitivity 42.4.2 Phantom images
52.4.3 Mouse images 5
3 Results 63.1 Performance of the PET detector modules outside
the MRI 63.2 Characterization of the cross-compatibility of PET
detector module and MRI 73.3 Phantom and small animal imaging of
the full ring PET 8
3.3.1 Spatial resolution and sensitivity 83.3.2 Phantom images
83.3.3 Mouse images 9
4 Discussion 9
5 Conclusion 11
1 Introduction
Positron emission tomography (PET) has attracted considerable
interest for the non-invasive visu-alization of small animals for
various preclinical studies. Currently, combined positron
emissiontomography and computed tomography (PET-CT) dedicated to
small animal imaging is commer-cially available [15] and has proven
to be a valuable imaging tool providing a fused image of
highresolution anatomical and quantitative functional information.
Moreover, combined PET and mag-netic resonance imaging (PET-MRI)
has been proposed for simultaneous functional and morpho-logical
images [68]. Extensive studies have been carried out by several
research groups to developMR-compatible PET based on
photomultiplier tube (PMT) using optical fiber technology [912],and
avalanche photodiodes (APDs) using RF shielding technology [13,
14].
Recently, the next generation photosensor Geiger-mode avalanche
photodiodes (GAPDs) [15],also called a solid state photomultiplier
(SSPM) [16], silicon photomultiplier (SiPM) [17], multipixel photon
counters (MPPC) [18] and micro-pixel avalanche photodiode (MAPD)
[19], was de-veloped. GAPDs consist of a densely packed matrix with
many microcells (100010000) rang-ing from 55 to 100100 m2 size, and
each microcell operates independently in a Geiger mode
1
-
2011 JINST 6 P08012
as on/off switch for the photons. The amplitude of output pulse
is proportional to the total numberof fired microcells, reflecting
the number of absorbed photons. GAPDs have been studied activelyas
a PET photosensor owing to their several key properties [20].
Compared to PMT, GAPDs havecompactness and insensitivity to high
magnetic fields. This allows GAPD to be located inside theMR bore
without the need for optical fibers to integrate hybrid PET-MRI. In
contrast to APDs,GAPDs have high gain (106) and low excess noise
factor (1.1), allowing operation with a sim-ple preamplifier. Their
fast response time (< 1 ns) allows a high true-to-random ratio
as well asthe development of a hybrid PET-MRI with time-of-flight
(TOF) capability in human whole-bodyapplication. Further advantages
over PMT and APDs are the low operating voltage (< 100 V)and
high uniformity (< 20%) among the pixels. In addition, the
fabrication costs can be reducedsubstantially because it can be
manufactured using a standard Metal Oxide Semiconductor
(MOS)production process [21].
The utility of the charge signal transmission approach, which
relays the charge signal fromthe photosensor to the remotely
located preamplifiers for PET signal transmission, was
recentlyreported [22]. This detector concept has several potential
merits because it allows the placementof amplifier units at a safe
distance for integrated PET-MR scanner, can decrease the space
re-quirements to insert a PET scanner into the restricted MR bore,
minimize the mutual interferencebetween PET and MRI, and eliminate
the need for placing RF shielding materials close to fieldof view
of the MR scanner. Moreover, it can reduce the deterioration by
temperature-related per-formance changes, which result from the
local heat production generated in the amplifier unit, inthe PET
system based on semiconductor photosensor. A previous study
verified that there was noconsiderable degradation in PET detector
performance, such as photopeak position of the 511 keV,energy
resolution and time resolution, even though the PET charge signal
was transferred via longcables (300 cm). On the other hand, the
scope of our previous study was limited to the
performancecharacterization of the PET detector that did not use
the channel reduction circuits to acquire to-mographic image from a
PET scanner.
The aim of this study was to develop a small animal PET scanner
based on GAPDs and thecharge signal transmission approach. The
performance of the PET detector modules was evaluatedand the
cross-compatibility between the PET detector module and MRI was
assessed. Quantitativeanalysis of the full ring prototype PET was
performed and tumor mouse images were acquired.
2 Materials and methods
2.1 System description
The full-ring PET contained 16 detector modules arranged in a
ring, 70 mm in diameter with anaxial field of view (FOV) of 13 mm
(figure 1). Each PET detector was comprised of a 44 lutetiumyttrium
oxyorthosilicate array (LYSO array, Sinocera, Shanghai, China) with
an individual crystalsize of 3 3 10-mm, arranged with a 3.3 mm
pitch. All crystal elements were polished andseparated with white
epoxy except for the photosensor face. The crystal block was
coupled directlyto a 3-side buttable GAPD array (SPMArray2, SensL,
Cork, Ireland). Each pixel of the GAPDarrays had a 2.85 2.85-mm
active area and 3,640 microcells of the 35 35-m. The feasibilityof
these GAPDs for the development of PET has been reported elsewhere
[2326].
2
-
2011 JINST 6 P08012
(a) (b)
Figure 1. Schematic diagram (a) and electronic components (b) of
small animal PET. (a): The animal PETcontained 16 LYSO and GAPD
detector modules arranged in a 70 mm diameter ring with an axial
field ofview (FOV) of 13 mm. (b): The GAPD outputs were transmitted
to the amplifier units and further processedin a PDC multiplexing
64 GAPD pixel output to one analog pulse signal and 6 bit position
information.
The 3-meter flexible flat cables (FFCs, New Grand TECH,
Shenzhen, China) were used totransmit the charge signal from the
PET detector to remotely located amplifier units, which con-sisted
of a 16 channel trans-impedance preamplifier (TIA) and a bias
regulation circuit.
The TIA converted the GAPD outputs to differential voltage
signals providing high gain(103) without introducing additional
noise and temperature-related gain instability. The biasregulation
circuit could adjust the bias voltage (28.532.5 V) finely at 8 mV
intervals using aprogrammable digital potentiometer (AD5231, Analog
Devices, MA, USA) and it was possible toprovide optimal operating
conditions for the GAPDs.
The position decoder circuits (PDC) [27] that were capable of
multiplexing the 64 GAPDpixel output (4 PET detectors 16
channels/detector) to one analog pulse signal and 6 bit
positioninformation were used. In addition, the PDC contained the
gain adjustable circuits to achieve gainhomogeneity for all
channels of the PET detectors. The 4 PDCs were used to multiplex
all the 256electrical PET signals (16 PET detectors 16
channels/detector), which simplified the PET systemdesign by
reducing the required ADC number and analog output lines from 256
to 4 channels.Four PDC output signals were fed into the data
acquisition (DAQ) unit using 10-meter twist-pairedcables and
co-axial cables to process the interaction position and analog
signal, respectively.
The DAQ unit (Lyrtech, Quebec, Canada) consisted of free-running
analog to digital convert-ers (ADC) and a field programmable gate
array (FPGA). The analog signals of the PDCs weredigitized at a 105
MHz sampling rate and a 14-bit vertical resolution in the
-1.25+1.25 V range.The digitized signals were processed further by
FPGA to calculate the accurate energy and timeinformation [28].
After signal processing, the output data containing the pulse
energy, arrival timeand position information were recorded in the
128 MB RAM in list mode format.
2.2 Performance measurement of the PET detector modules outside
MRI
A 200-kBq 22Na point source placed centrally between the paired
PET detectors was used to ir-radiate the LYSO-GAPD arrays. The
energy and time spectra were acquired at room temperaturewithout
additional cooling of the PET detectors. The energy and time
resolution were calculatedas the full width at half maximum (FWHM)
of the Gaussian distribution plot. The lower energy
3
-
2011 JINST 6 P08012
threshold was set to 350 keV and the coincidence time window was
4 ns. The variation of thecount uniformity and photopeak position
of the 16 channels flood histogram were calculated as theratio of
the standard deviation to the average value.
2.3 Characterization of the cross-compatibility of PET detector
module and MRI
The performance measurements of the PET detectors were also
repeated in 7-T MRI (BrukerBioSpec, Ettlingen, Germany) to
characterize the effect of MRI on the PET detector module. Apair of
LYSO-GAPD PET detectors was inserted inside the MRI bore between
the RF coil andgradient coils. The GAPD outputs were transmitted to
the amplifier units using 3-meter FFCs. ThePET electronics were
positioned outside of the 5-Gauss line (1.5-meters away from the
magnetisocenter) in the MR to minimize the mutual interference
between PET and MRI. In this study, noelectromagnetic shielding was
introduced to protect the PET components from the MR gradientand RF
field.
A uniform cylindrical phantom (30 mm diameter, 100 mm length)
filled with a copper sul-fate solution was imaged to examine the
effect of the PET detector modules on MR images. TheCuSO4-filled
phantom was placed isocentrically inside the RF-coil, and the
transaxial images wereacquired with and without a pair of LYSO-GAPD
PET detectors inside the MR bore. The standardMR imaging sequences,
including the Gradient echo (TR = 205 ms, TE = 6 ms, FA = 15
degree),Spin echo T1 (TR = 419 ms, TE = 8 ms) and Spin echo T2 (TR
= 3,000 ms, TE = 75 ms) wereused in this study. The ParaVision
software (Bruker BioSpec, Ettlingen, Germany) was used toacquire
and process the MR data, such as data acquisition, analysis,
reconstruction and visualiza-tion. Representative three transaxial
MR image slices with a 5 mm interval in the axial directionwere
analyzed quantitatively. A region of interest (ROI) was drawn at
the center of the phantomimage enclosing approximately 80% of the
phantom and the uniformity and signal to noise ratio(SNR) were
calculated from the ROI. The experimental tests were repeated five
times to minimizethe measurement errors. In each measurement, the
MR phantom was re-positioned and the RF-coilwas re-inserted and
re-tuned.
2.4 Phantom and small animal imaging of the full ring PET
The list-mode data of the full ring PET were rebinned using a
single slice rebinning (SSRB) methodand the valid events were
sorted into the 2D sinogram, which had 43 samples in the transverse
di-rection and 79 angular samples. The transverse sampling distance
was 1.5 mm near the CFOV.The sinogram was normalized for the
detector efficiency, and these normalization factors were
es-timated from a direct inversion of the sinogram acquired with a
uniform cylindrical phantom filledwith a 18F-FDG solution. The
missing data caused by the effect of the gaps was compensatedfor by
a nearest neighbor 1-D interpolation in the radial direction [29].
The PET images were re-constructed by a 2D filtered backprojection
(2D FBP) using a Hanning filter with a cutoff at theNyquist
frequency.
2.4.1 Spatial resolution and sensitivity
The spatial resolution of the prototype PET was measured using a
glass capillary tube with aninner diameter of 0.5 mm. The line
source was placed at five radial offset locations of 0, 10, 15,
4
-
2011 JINST 6 P08012
Figure 2. Experiment setup for the mouse imaging study. The
prototype PET was installed with a heatingpad, anesthesia and
respiration monitoring systems for the in-vivo study.
20 and 25 mm in the FOV. Each line source was filled with
approximately 370-kBq of a 18F-FDGsolution. The radial profiles of
each location were fitted with Gaussian profiles and its width
wasmeasured by the FWHM. The coincidence sensitivity of the
prototype PET was measured using apoint source with an inner
diameter of 0.5 mm. A 200-kBq 22Na point source was located
preciselyat the center of the ring and the PET data was acquired
for 1 min. The data recorded in list modeformat was sorted using
four different energy windows ranging from 10% (460560 keV) to40%
(300710 keV). A coincidence time window of 4 ns was applied.
The system sensitivity was calculated as the number of detected
events divided by the numberof decays by positron emissions that
were predicted to have occurred during the acquisition period.
2.4.2 Phantom images
Two cable lengths (10 cm and 300 cm FFCs), connecting the GAPD
arrays to the amplifier units,were used to evaluate the effect of
the cable length on the PET image. Custom-made hot-spherephantoms,
60 mm in diameter, were used to examine the imaging performance.
The sphere diame-ters were 3, 4, 5, 6 and 7 mm and the
center-to-center distance between the spheres was twice
theirdiameter. The phantoms were filled with a 20 MBq 18F-FDG
solution. The PET imaging data wasacquired for 10 minutes.
2.4.3 Mouse images
An in vivo rodent imaging study was performed in accordance with
the protocols approved bythe Samsung Biomedical Research Institute
(SBRI) in Korea. A male mouse with a tumor inthe right thigh was
injected with 100-MBq of 18F-FDG through the tail vein and imaged
after 1hour of radiotracer uptake. The mouse was placed on a carbon
animal bed with a heating pad thatmaintained a temperature of 37 C.
During the PET scan, the mouse was anesthetized by
isofluraneinhalation and respiration was monitored. The PET imaging
data was acquired for 10 minutes/bedat 2 different bed positions.
Figure 2 shows the experimental setup.
5
-
2011 JINST 6 P08012
(a) (b)
Figure 3. Representative energy spectrum (a) and time spectrum
(b) acquired outside MRI. The energy andtime resolution of the PET
detector were 16.4% and 1.5 ns, respectively.
(a) (b)
Figure 4. Flood histogram (a) and plots of the energy resolution
and photopeak position (b) of the 4 4LYSO and GAPD array. The
changes in the count uniformity in the flood histogram and the
photopeakposition were 3.3% and 3.9%, respectively.
3 Results
3.1 Performance of the PET detector modules outside the MRI
Figure 3 shows the representative energy and time spectrum of
the PET detector. The mean energyresolution was 17.6%, ranging from
16% to 19%. The time resolution obtained was 1.5 ns for 511-keV
photons. Figure 4.a shows the flood histogram of the data acquired
from the PET detector.The variation of the count uniformity was
3.3%. The variation of the photopeak position of theLYSO-GAPD
detector was 3.9% and was improved after gain adjustment of the PDC
(figure 4.b).
6
-
2011 JINST 6 P08012
(a) (b)
Figure 5. Energy spectra (a) and Time spectra (b) acquired both
outside and inside the MRI. No obviousperformance degradation of
the PET detectors was observed.
Figure 6. Quantitative analysis of the MR phantom images in
terms of the SNR (left) and uniformity (right):gradient echo (first
row), T1 weighted spin echo (second row) and T2 weighted spin echo
(third row).
3.2 Characterization of the cross-compatibility of PET detector
module and MRI
The energy and time spectra were acquired simultaneously for
approximately 5 min, whereas MRimaging was performed using three
different sequences, as shown in figure 5. No obvious per-formance
degradation of the PET detectors was observed, as measured by the
energy and timeresolution. In addition, the photopeak position and
coincidence count rate were similar regardlessof whether the PET
detectors had been operated outside or inside the MRI.
7
-
2011 JINST 6 P08012
Figure 7. Spatial resolution of the developed PET. The measured
radial resolution ranged from 2.8 mm to4.1 mm.
Table 1. Sensitivity at the CFOV for the 4 different energy
window settings.
Energy window Total counts Sensitivity (%)
460560 keV (10%) 11,396 0.29
410615 keV (20%) 18,894 0.48
350650 keV (30%) 25,843 0.65
300710 keV (40%) 31,050 0.78
There were no significant artifacts or distortions observed in
the MR phantom images. Figure 6shows the calculated values of the
uniformity and SNR for three MR sequences. There was nomajor loss
caused by inserting the PET detector modules in the SNR and
uniformity of MR images.
3.3 Phantom and small animal imaging of the full ring PET
3.3.1 Spatial resolution and sensitivity
Figure 7 shows the measured phantom images and radial profiles
for 5 different radial offsets acrossthe useful FOV. The radial
resolution was 2.8 mm FWHM at the CFOV, which increased to 4.1
mmwith a 20 mm offset.
Table 1 lists the sensitivity at the CFOV for different energy
window settings. The sensitivitywas corrected for a branching ratio
of 22Na (0.906). The system sensitivity increased approximatelyfour
times as the energy window was changed from 460560 keV to 300710
keV. The prototypePET had a peak sensitivity of 0.65% in the
standard energy window of 350650 keV.
3.3.2 Phantom images
Figure 8 shows transverse phantom images and line profiles
acquired using the 10 cm (a) and300 cm (b) FFCs connecting GAPDs
and preamplifiers. The 3 mm diameter was clearly resolvedin the
hot-sphere phantom images. As expected from previous studies [22],
there was no degrada-tion of the PET image quality caused by
employing a long cable (300 cm) from the GAPDs to thepreamplifiers
used for PET signal transmission.
8
-
2011 JINST 6 P08012
(a) (b)
Figure 8. Hot-sphere phantom images and line profiles were
acquired using the 10 cm (a) and 300 cm (b)FFCs between GAPDs and
preamplifiers. There was no degradation of the PET image quality
caused by thelong cable (300 cm) from the GAPDs to the
preamplifiers in the PET signal transmission.
3.3.3 Mouse images
A PET mouse image was acquired to demonstrate the potential of
this prototype system for an invivo study. Figure 9 shows the
reconstructed transaxial PET images of a mouse. The mouse brainand
tumor displaying the glucose metabolism patterns were well
imaged.
4 Discussion
A small animal PET based on the GAPDs and a charge signal
transmission approach was devel-oped. In this system, the charge
signals of GAPDs were transmitted to the amplifier units
positionedfar away from the PET detectors or outside the MR bore
using long transmission cables. The initialPET images of the
phantom and tumor mouse were obtained in this study. In addition,
simultane-ous PET and MR data was acquired with no deterioration in
the performance of both PET detectormodules and MRI. The concept
proposed in this study might provide several technical
advantages.
It is feasible to develop a PET system based on a semiconductor
photosensor with no obviousloss of PET performance. Unfortunately,
APDs have not achieved the traditional PMT perfor-mance. The low
internal gain of the APD (< 102) requires the use of
sophisticated low noisepreamplifiers [30]. The overall time
resolution of the imaging device is reduced by the signifi-cantly
inferior timing properties of the APDs due to the slow rise time,
high time jitter, high noisefactor, and low signal-to-noise ratios
[21]. This results in a wider coincidence time window and an
9
-
2011 JINST 6 P08012
Figure 9. In vivo study of the prototype PET. The PET imaging
data was acquired for 10 min at the differentbed positions and the
mouse brain (top) and tumor (bottom) displaying the glucose
metabolism patterns werewell imaged.
increased number of random coincidence counts, which will
degrade the image quality of PET. Onthe other hand, GAPD can
overcome these technical challenges. Simple electronics can be used
inthe development of PET based on GAPD. In addition, GAPD can
provide good PET performance,such as energy resolution (figure 3.
(a)), time resolution (figure 3. (b)), count uniformity (figure
4.(a)) and long-term stability [25]. The inherent characteristics
of the GAPDs could improve the PETimage performance by providing a
high SNR and high true-to-random coincidence rates.
In addition, the energy resolution of the 511-keV photopeaks was
17% on average for allcrystals, which is better than the 25% [31]
and 27% [32] energy resolutions obtained for pre-viously reported
animal PET scanners based on GAPD and LGSO. This may be the result
of thedetector configuration, where the individual LYSO crystal was
coupled one to one to a separatepixel of the GAPD and the output
signals from each PET detector module were processed inde-pendently
using the PDC, which could eliminate the performance degradation
caused by opticalcrosstalk through the light guide inserted between
the crystal block and photosensor array in theconventional detector
configuration based on the Anger type channel reduction
circuit.
As shown in figure 8, the developed PET system using 300 cm long
cables produced phantomimages without noticeable quality
degradation. Moreover, the charge signal transmission approachcould
minimize the deterioration by temperature changes. A potential
technical hurdle is that thePET performance degrades gradually by
the heat generated from the amplifiers unit in the PET sys-tem
based on semiconductor photosensor [22]. Previous studies using a
semiconductor photosensor
10
-
2011 JINST 6 P08012
employed a costly cooling system or gain control system to
minimize the temperature-related per-formance variations of the
photosensor [3336]. This detector concept, photosensors and
amplifiersfabricated in the separated housing box, could minimize
the temperature-related performance vari-ations of the photosensor,
such as the gain, PDE and changes in dark currents, which could
affectthe PET performance, such as altered photo-peak position,
decreased count rate, degraded energyresolution and time
resolution.
In terms of the MR compatibility, the developed PET system only
requires the placement ofnon-magnetic PET components, scintillation
crystals and GAPD arrays without preamplifiers andsubsequent
electrical circuits, inside the MR bore. The PET system requires
only low DC powersupplies with low current consumption (32 V, <
2 mA) inside the MR bore. Moreover, this systemdoes not introduce
conducting materials, such as RF shielding materials and electronic
circuitsinside the MR bore, which can produce eddy currents and
heat from the high power switchinggradient coil [3739]. The free
scalability of the PET geometry, such as transaxial and axial
FOV,are an additional advantage by locating the amplifier units at
low fringe field areas, which coulddecrease the space requirements
to insert a PET scanner into the restricted MR bore.
This study had some limitations. First, the spatial resolution
was relatively poor compared toother systems based on the Anger
logic circuit, which was caused by the one to one coupling ofan
individual crystal with a separate pixel of large area GAPDs (
33mm2). Although the PETspatial resolution was poor, these results
demonstrate that it was feasible to develop small animalPET using
GAPDs and charge signal transmission approach technologies.
Moreover, the coinci-dence detection efficiency was improved by a
factor of 2 when the crystal length was increasedfrom 5 mm to 10 mm
at the expense of parallax error. A sensitivity of 0.7% was
acquired usingthe developed PET, even though it had a shorter axial
FOV and larger transaxial FOV than otherprototype systems [12-14,
40]. Second, simultaneous PET-MR images were not acquired becauseit
was difficult to use a radioisotope in a MR imaging site. On the
other hand, the feasibility ofhybrid PET-MRI with this system
design was observed to some extent (figure 5 and 6). Neverthe-less,
further studies will be needed to improve the spatial resolution
using smaller size GAPDs andacquire simultaneous PET-MR images
using a range of MR imaging sequences.
5 Conclusion
A small animal PET was developed based on GAPDs and the charge
signal transmission approach.The PET detector performance and
cross-compatibility were examined. Quantitative analysis ofthe full
ring prototype PET was performed and the tumor mouse images were
acquired success-fully. These results demonstrated that the GAPD
and charge signal transmission approach allowthe development of
high performance small animal PET with improved MR
compatibility.
Acknowledgments
This research was supported by the Converging Research Center
Program through the National Re-search Foundation of Korea (NRF)
funded by the Ministry of Education, Science and
Technology(2011K000715), and by the Technology Innovation Program
funded by the Ministry of KnowledgeEconomy (10030029), Republic of
Korea.
11
-
2011 JINST 6 P08012
References
[1] A. Chatziioannou et al., Detector development for microPET
II: a 1 l resolution PET scanner forsmall animal imaging, Phys.
Med. Biol. 46 (2001) 2899.
[2] Y.C. Tai et al., MicroPET II: design, development and
initial performance of an improved microPETscanner for small-animal
imaging, Phys. Med. Biol. 46 (2003) 1519.
[3] Y.C. Wang et al., Performance evaluation of the GE
healthcare eXplore VISTA dual-ringsmall-animal PET scanner, J.
Nucl. Med. 47 (2006) 1891.
[4] M.C. Huisman et al., Performance evaluation of the Philips
MOSAIC small animal PET scanner, Eur.J. Nucl. Med. 34 (2007)
532.
[5] C.C. Constantinescu et al., Performance evaluation of an
Inveon PET preclinical scanner, Phys. Med.Biol. 54 (2009) 2885.
[6] H. Zaidi et al., Simultaneous PET/MR will replace PET/CT as
the molecular multimodality imagingplatform of choice, Med. Phys.
34 (2007) 1525.
[7] B.J. Pichler et al., Latest advances in molecular imaging
instrumentation, J. Nucl. Med. 49 (2008) 5S.
[8] H.F. Wehrl et al., Pre-clinical PET/MR: technological
advances and new perspectives in biomedicalresearch, Eur. J. Nucl.
Med. 36 (2009) 56.
[9] R.B. Slates et al., A study of artefacts in simultaneous PET
and MR imaging using a prototype MRcompatible PET scanner, Phys.
Med. Biol. 44 (1999) 2015.
[10] P.K. Marsden et al., Simultaneous PET and NMR, Brit. J
Radiol. 75 (2002) S53.
[11] R.R. Raylman et al., Simultaneous MRI and PET imaging of a
rat brain, Phys. Med. Biol. 51 (2006)6371.
[12] S. Yamamoto et al., A multi-slice dual layer MR-compatible
animal PET system, IEEE Trans. Nucl.Sci. 56 (2009) 2706.
[13] C. Woody et al., Preliminary studies of a simultaneous
PET/MRI scanner based on the RatCAP smallanimal tomograph, Nucl.
Instrum. Meth. 571 (2007) 102.
[14] M.S. Judenhofer et al., Simultaneous PET-MRI: a new
approach for functional and morphologicalimaging, Nature Medicine
14 (2008) 459.
[15] I. Britvitch et al., Characterisation of Geiger-mode
avalanche photodiodes for medical imagingapplications, Nucl.
Instrum. Meth. 571 (2007) 308.
[16] M. McClish et al., Characterization and scintillation
studies of a solid-state photomultiplier, Nucl.Instrum. Meth. 572
(2007) 1065.
[17] D.J. Herbert et al., The silicon photomultiplier for
application to high-resolution Positron EmissionTomography, Nucl.
Instrum. Meth. 573 (2007) 84.
[18] T. Takeshita, Development of MPPC, Pramana 69 (2007)
1079.
[19] Z. Sadygov et al., Three advanced designs of micro-pixel
avalanche photodiodes: their present status,maximum possibilities
and limitations, Nucl. Instrum. Meth. 567 (2006) 70.
[20] D. Renker, Geiger-mode avalanche photodiodes, history,
properties and problems, Nucl. Instrum.Meth. 567 (2006) 48.
[21] R. Lecomte, Novel detector technology for clinical PET,
Eur. J. Nucl. Med. Mol. I. 36 (2009) 69.
12
http://dx.doi.org/10.1088/0031-9155/46/11/310http://dx.doi.org/10.1088/0031-9155/48/11/303http://dx.doi.org/10.1007/s00259-006-0271-7http://dx.doi.org/10.1007/s00259-006-0271-7http://dx.doi.org/10.1088/0031-9155/54/9/020http://dx.doi.org/10.1088/0031-9155/54/9/020http://dx.doi.org/10.1118/1.2732493http://dx.doi.org/10.2967/jnumed.108.045880http://dx.doi.org/10.1007/s00259-009-1078-0http://dx.doi.org/10.1088/0031-9155/44/8/312http://dx.doi.org/10.1088/0031-9155/51/24/006http://dx.doi.org/10.1088/0031-9155/51/24/006http://dx.doi.org/10.1109/TNS.2009.2030192http://dx.doi.org/10.1109/TNS.2009.2030192http://dx.doi.org/10.1016/j.nima.2006.10.039http://dx.doi.org/10.1038/nm1700http://dx.doi.org/10.1016/j.nima.2006.10.089http://dx.doi.org/10.1016/j.nima.2007.01.001http://dx.doi.org/10.1016/j.nima.2007.01.001http://dx.doi.org/10.1016/j.nima.2006.11.020http://dx.doi.org/10.1016/j.nima.2006.05.215http://dx.doi.org/10.1016/j.nima.2006.05.060http://dx.doi.org/10.1016/j.nima.2006.05.060http://dx.doi.org/10.1007/s00259-008-1054-0
-
2011 JINST 6 P08012
[22] J. Kang et al., A feasibility study of photosensor charge
signal transmission to preamplifier using longcable for development
of hybrid PET-MRI, Med. Phys. 37 (2010) 5655.
[23] D.R. Schaart et al., A novel, SiPM-array-based, monolithic
scintillator detector for PET, Phys. Med.Biol. 54 (2009) 3501.
[24] P.D. Olcott et al., Novel electro-optical coupling
technique for magnetic resonance-compatiblePositron Emission
Tomography detectors, Mol. Imaging 8 (2009) 77.
[25] K.J. Hong et al., Performance evaluation of a PET detector
consisting of an LYSO array coupled to a44 array of large-size GAPD
for MR compatible imaging, 2011 JINST 6 P05012.
[26] J. Kang et al., A dual-ended readout PET detector module
based on GAPDs with large-areamicrocells, 2011 JINST 6 P07003.
[27] J.H. Jung et al., Development of a position decoder circuit
for PET consisting of GAPD arrays, Nucl.Instrum. Meth. 621 (2010)
310.
[28] W. Hu et al., A simple and improved digital timing method
for positron emission tomography, Nucl.Instrum. Meth. 622 (2010)
219.
[29] S. Shokouhi et al., System performance simulations of the
RatCAP awake rat brain scanner, IEEETrans. Nucl. Sci. 52 (2005)
1305.
[30] B.J. Pichler et al., Integrated low-noise low-power fast
charge-sensitive preamplifier for avalanchephotodiodes in JFET-CMOS
technology, IEEE Trans. Nucl. Sci. 46 (2001) 2370.
[31] S.I. Kwon et al., Development of small-animal PET prototype
using silicon photomultiplier (SiPM):initial results of phantom and
animal imaging studies, J. Nucl. Med. 52 (2011) 572.
[32] S. Yamamoto et al., Development of a Si-PM-based
high-resolution PET system for small animals,Phys. Med. Biol. 55
(2010) 5817.
[33] C. Catana et al., Simultaneous in vivo positron emission
tomography and magnetic resonanceimaging, Proc. Natl. Acad. Sci.
U.S.A. 105 (2008) 3705.
[34] A. Kolb et al., Evaluation of Geiger-mode APDs for PET
block detector designs, Phys. Med. Biol. 55(2010) 1815.
[35] H. Herzog et al., The current state, challenges and
perspectives of MR-PET, NeuroImage 49 (2010)2072.
[36] S. Yamamoto et al., A temperature-dependent gain control
system for improving the stability ofSi-PM-based PET systems, Phys.
Med. Biol. 56 (2011) 2873.
[37] C.R. Camacho et al., Nonsusceptibility artifacts due to
metallic objects in MR-imaging, J. Magn.Reson. Imaging 5 (1995)
75.
[38] H. Graf et al., Effects on MRI due to altered rf
polarization near conductive implants or instruments,Med. Phys. 33
(2006) 124.
[39] B.J. Pichler et al., Positron Emission Tomography/Magnetic
Resonance imaging: the next generationof multimodality imaging?,
Semin. Nucl. Med. 38 (2008) 199.
[40] Y.B. Wu et al., PET performance evaluation of an
MR-compatible PET insert, IEEE Trans. Nucl. Sci.56 (2009) 574.
13
http://dx.doi.org/10.1118/1.3495683http://dx.doi.org/10.1088/0031-9155/54/11/015http://dx.doi.org/10.1088/0031-9155/54/11/015http://dx.doi.org/10.1088/1748-0221/6/05/P05012http://dx.doi.org/10.1088/1748-0221/6/07/P07003http://dx.doi.org/10.1016/j.nima.2010.04.028http://dx.doi.org/10.1016/j.nima.2010.04.028http://dx.doi.org/10.1016/j.nima.2010.05.047http://dx.doi.org/10.1016/j.nima.2010.05.047http://dx.doi.org/10.1109/TNS.2005.858236http://dx.doi.org/10.1109/TNS.2005.858236http://dx.doi.org/10.1109/23.983270http://dx.doi.org/10.2967/jnumed.110.079707http://dx.doi.org/10.1088/0031-9155/55/19/013http://dx.doi.org/10.1073/pnas.0711622105http://dx.doi.org/10.1088/0031-9155/55/7/003http://dx.doi.org/10.1088/0031-9155/55/7/003http://dx.doi.org/10.1016/j.neuroimage.2009.10.036http://dx.doi.org/10.1016/j.neuroimage.2009.10.036http://dx.doi.org/10.1088/0031-9155/56/9/015http://dx.doi.org/10.1002/jmri.1880050115http://dx.doi.org/10.1002/jmri.1880050115http://dx.doi.org/10.1118/1.2132571http:/dx.doi.org/10.1053/j.semnuclmed.2008.02.001http://dx.doi.org/10.1109/TNS.2009.2015448http://dx.doi.org/10.1109/TNS.2009.2015448
IntroductionMaterials and methodsSystem descriptionPerformance
measurement of the PET detector modules outside MRICharacterization
of the cross-compatibility of PET detector module and MRIPhantom
and small animal imaging of the full ring PETSpatial resolution and
sensitivityPhantom imagesMouse images
ResultsPerformance of the PET detector modules outside the
MRICharacterization of the cross-compatibility of PET detector
module and MRI Phantom and small animal imaging of the full ring
PETSpatial resolution and sensitivityPhantom imagesMouse images
DiscussionConclusion