jPET Project Team Abstracts in
IEEE NSS/MIC 2006
Thursday, Nov. 2
M06: MIC Poster 1
10:30-12:00, Atlas Ballroom
|
Session Chairs: |
Kenneth Wong, Georgetown University |
M06-77:
Inter-Crystal Scatter Identification for a Depth-Sensitive Detector Using
Multi-Anode Outputs
1National
Institute of Radiological Sciences, Chiba, Japan
2Shimadzu Corporation, Kyoto, Japan
3Tokyo Metropolitan Institute of Gerontolog,
Tokyo, Japan
In conventional PET detectors, detected events are projected to 2D position histogram using the Anger calculation for crystal identification. Inter-crystal scatter (ICS) events causes mispositioned crystals because peaks with projected each crystal in the histogram are blurred. A depth of interaction (DOI) detector has been developed for the small animal PET scanner: jPET-RD. This DOI detector uses 4-layered arrays of 32 x 32 crystals and a 256-channel multi-anode flat panel photomultiplier tube (FP-PMT). Each crystal element is 1.45 mm x 1.45 mm x 4.5 mm. The FP-PMT has a large detective area and small anode pitch. We think that the FP-PMT has a potential for the trace of the scattered gamma rays in the crystals. In this study, we therefore propose a novel method for ICS estimation using the principle component analysis (PCA). The PCA is applied to multiple-anode outputs in order to discriminate photoelectric events from ICS events. The 1st principle component depends on number of anode outputs deposited energy through the preprocessing. Numerical simulation results shows that the ratio of the ICS identification by the proposed method is about 73 % at a 511 keV uniform irradiation. Also, PCA method can be identified only events which have large effect on spatial resolution in the ICS events. The proposed method can archive a true subtraction of ICS from measured events.
M06-107:
A Healthy Volunteer FDG-PET Study on the Limit of the Spatial Resolution due to
Annihilation Radiation Non-Collinearity
K. Shibuya1, E. Yoshida1,
F. Nishikido1, T. Suzuki2,
1Department
of Biophysics, Moleculer Imaging Center, National
Institute of Radiological Sciences,
2Department of Dose Assessment,
We provide a new viewpoint for a fundamental but little investigated problem in positron emission tomography (PET): non-collinearity of annihilation radiation from the human body. The cause of the small angular deviation from 180° is described as well as how to evaluate it under a spatially distributed radiation source and a limited acquisition time. An elegant conversion from the photopeak spectrum into the angular distribution is done based on the conservation laws of momentum and energy to avoid problems in the direct coincidental measurement. A healthy volunteer study using 18FDG and a Ge semiconductor detector reveals the distribution as a Gaussian function with the FWHM of 0.54°, which is about 15% larger than the value reported for pure water. From the result, we can calculate the physical limit of PET spatial resolution by convoluting the angular deviation with the positron range. For example, the FWHM of the line spread function is 0.5 mm for the detectors ring diameter of 10 cm, e.g., small animal PET, 1.2 mm for that of 40 cm, e.g., human brain PET, and 2.1 mm for that of 80 cm, e.g., human wholebody PET.
Friday, Nov. 3
M11: MIC Poster 2
13:30-15:00, Atlas Ballroom
|
Session Chairs: |
Philippe Després, Physics Research
Laboratory, University of California, San Francisco |
M11-95:
Multi-Channel Waveform Sampling ASIC for Animal PET System
K. Shimazoe1, Y. J. Yoel2, H. Takahashi1, T. Kojo3, Y. Minamikawa2, K. Fujita2, H. Murayama4
1Dept.
Bioengineering, The University of Tokyo, TOKYO, JAPAN
2Dept. Quantum Engineering and System Science, The University of
Tokyo, TOKYO, JAPAN
3Dept. Nuclear Engineering and Management, The University of Tokyo,
TOKYO, JAPAN
4National Institute of Radiological Sciences, Chiba, Japan
We have designed and fabricated 2.4mm by 2.4mm 2ch 100MHz/6bits waveform sampling front-end (WSFE) ASIC for PET(Positron Emission Tomography) with multiplexer readouts. This chip was designed for GSO-APD gamma-ray detectors and has the function of waveform sampling” at the speed of 100MHz.This chip consists of charge-sensitive preamplifier, VGA (Variable Gain Amplifier), folding-ADC and digital readout circuit and converts the waveform of detector signal to digital values at the speed of ~100MHz. Digitization in early stage in one chip enables flexible digital signal processing to acquire timing and energy information of gamma-rays which is indispensable for PET system. Digital circuit has three ways to read signals out, parallel readout, 32words FIFO memory and 2 to 1 multiplexer. The 2 to 1 multiplexer reduces the number of ASIC pins half by using fast clock with double frequency and enables to include more channels in one chip. We designed a new WSFE ASIC which includes parallel readout, 2 to 1 multiplexers, 3 to1 multiplexers and 6 to 1 multiplexers. These multiplexers will greatly decrease the number of pins and realize high integration. This chip will be used for APD-based DOI PET system which requires multi-channel ASIC.
M11-146:
The jPET-D4: Performance Evaluation of Four-Layer DOI-PET Scanner Using the
NEMA NU2-2001 Standard
1National
Institute of Radiological Sciences, Chiba, Japan
2Tokyo Metropolitan University, Tokyo, Japan
3Hamamatsu Photonics K.K., Shizuoka, Japan
4Shimadzu Co., Kyoto, Japan
5Kitasato University, Kanagawa, Japan
We have developed a high-performance brain PET scanner, jPET-D4. This scanner is designed to achieve not only high spatial resolution but also high sensitivity using four-layered depth-of-interaction (DOI) detectors. The scanner has five block detector rings with the ring diameter of 390 mm and each block detector ring consists of 24 DOI detectors. In this paper, we present sensitivity and count-rate performance of the jPET-D4 using the NEMA NU2-2001 standard. The average scatter fraction for this system is 42 % with energy window of 400-600 keV. The sensitivity at center FOV is 19.5 kcps/MBq with energy window of 400-600 keV. Peak NECRs (NEC1R and NEC2R) with 10 ns coincidence time window are 82 kcps at 8.7 kBq/ml and 56 kcps at 7.6 kBq/ml, respectively. These first evaluation measurements promise the jPET-D4 has excellent performance.
M11-218:
Performance Evaluation of jPET-D4 with the
T. Hasegawa1,
1Allied
Health Sciences, Kitasato University, Kitasato, Sagamihara, Kanagawa, Japan
2Molecular Imaging center, National Institute of Radiological
Sciences, Inage, Chiba, Japan
3Human Health Sciences, Tokyo Metropolitan University, Arakawa,
Tokyo, Japan
4Science and Technology, Chiba University, Inage, Chiba, Japan
We have been carrying out performance evaluation of jPET-D4 (a new
brain PET scanner with a four-layer depth-of-interaction detector scheme) by
phantom measurement and
Saturday, Nov. 4
M14: MIC Poster 3
10:30-12:00, Atlas Ballroom
|
Session Chairs: |
Youngho Seo, |
M14-72:
A Monte Carlo Simulation Study on Detector Arrangement for a Small Bore DOI-PET
Scanner: jPET-RD
T. Kobayashi1, T. Yamaya2, H. Takahashi1, K. Kitamura3, T. Hasegawa4, H. Murayama2, M. Suga1
1Image
Science and Technology, Chiba university graduate school of Science and
technology, Chiba, Japan
2Molecular Imaging Center, The National Institute of Radiological
Sciences, Chiba, Japan
3Medical Systems Division, The Shimadzu Corporation, Kyoto, Japan
4School of Allied Health Sciences, Kitasato
University, Sagamihara, Japan
To develop a small bore DOI-PET scanner dedicated to small animals: jPET-RD, we performed Monte-Carlo simulations using the GATE based on Geant4, and investigated the influence of closing detectors to an imaging object upon the imaging and count rate performance. The jPET-RD is based on a large size depth of interaction (DOI) block detector which consists of 4-layered array of 32x32 LSO crystals (1.4mm x 1.4 mm x 4.5mm) and a 256ch flat panel position sensitive photomultiplier tube. In this work, three geometries based on the detectors are simulated: two rings of six detector blocks arranged in a hexagonal shape (85 mm diameter FOV), four detector blocks arranged in a tetragonal shape (49 mm) and in an overlapped tetragonal shape (38 mm). Reconstructed images shows that the hexagonal scanner has better spatial resolution and its uniformity than the other scanners, and the resolution performance of the tetragonal scanner declines due to the large inter-detector gaps; however it can be minimized by overlapping each detector. Additionally, count rate simulation results show that the smaller bore geometry can provide higher sensitivity because of its larger solid angle. While it clearly affects noise equivalent count rate (NECR) due to its high dead-time capability; however the parallel readout with appropriate anode segmentation can improve the NECR at 20MBq by factors of 1.1 to 1.4 compared with the case that 256ch anodes read out by one front-end circuit.
M14-96:
Spatial Resolution Measured by a Prototype System of Two 4-Layer DOI Detectors
for jPET-RD
F. Nishikido1, T. Tsuda1,
N. Inadama1,
1National
Institute of Radiological Sciences,
2Shimadzu Corporation,
We are developing a small animal PET scanner, “jPET-RD”
to achieve high sensitivity as well as high spatial resolution by the use of
four-layer depth of interaction (DOI) information of the detector. The jPET-RD is designed to have two detector rings composed of
six DOI detectors arranged hexagonally. The diameter of the field of vies is
8.8 mm, which is smaller than typical small animal PET scanners. Each detector
module consists of a 32 × 32 × 4 LYSO (Lu: 98%, Y: 2%) crystal array and a
256-channel flat panel position sensitive photomultiplier tube. The size of
each crystal element is 1.46 mm × 1.46 mm × 4.5 mm. The crystal block is placed
on the central area of a 256ch FP-PMT (49 mm × 49 mm useful area) and coupled
with silicone rubber.
In the previous study, we developed a prototype detector for the jPET-RD. In this work, we develop a prototype imaging
system for the jPET-RD. The system is composed of two
prototype detectors and electrical circuits. We evaluated the spatial
resolution of reconstructed images with the one pair system of the two 4-Layer
DOI detectors for jPET-RD. The spatial resolution of
1.5mm was obtained at the center of filed of view (FOV) by FBP algorithm. The
spatial resolutions of better than 2 mm in whole field of view were also
achieved.
M14-135:
Optimization of
N. Inadama1, H. Murayama1,
T. Tsuda1, F. Nishikido1, K. Shibuya1,
T. Yamaya1,
1Department
of Biophysics, Molecular Imaging Center, National Institute of Radiological
Sciences, Chiba, Japan
2Graduate School of Science and Technology, Chiba university, Chiba,
Japan
3Science and Engineering, Waseda
university, Tokyo, Japan
Previously, we proposed an 8-layer depth-of- interaction (DOI)
encoding method for a PET detector and proved its validity. The layer of
interaction is identified by hybrid method; the scintillation light control by
the original reflector arrangement for 4-layer DOI encoding and the pulse shape
discrimination for 2-layer DOI encoding. Four layers then consist of the
scintillator of different pulse shape from the scintillator for the other four
layers.
We investigated the effect of crystal arrangement difference on detector
performance to optimize the DOI detector. The two kind layers can be arranged
in two ways; stacked alternately or set in the upper and lower four layers
respectively. Since the two crystal arrangements are supposed to show different
detector performance, we measured it on each arrangement. The results show
better performance of the stacked arrangement in pulse shape discrimination,
while inferior in crystal identification among the same kind scintillation
crystals composing four layers. There was no particular difference between the
two crystal arrangements in light outputs and energy resolutions of each layer.
M14-390:
First Human Brain Images of the jPET-D4 Using 3D OS-EM
with a Pre-Computed System Matrix
T. Yamaya1,
1Molecular
Imaging Center, National Institute of Radiological Sciences, Chiba, JAPAN
2Shimadzu Co., Kyoto, Japan
3Interdisciplinary Graduate School of Science and Engineering, Tokyo
Institute of Technology, Yokohama, Japan
The jPET-D4 is a novel brain PET scanner which aims to achieve not only high spatial resolution but also high scanner sensitivity by measuring 4-layer depth-of-interaction (DOI) information. In this work, we present software strategies for 3D image reconstruction and imaging performance of the jPET-D4 prototype. The dimensions of a system matrix for the jPET-D4 become 4 billion (coincidence pairs) x 5 million (image elements) when a 25cm diameter FOV is sampled by a 1.5mm3 voxel. The size of the system matrix is estimated at 142peta (P) byte with the accuracy of 8 byte per element. The on-the-fly calculation is usually used to deal with a huge system matrix. However we can not avoid the extension of calculation time when we improve the accuracy of system modeling. In this work, we proposed an alternative approach based on the pre-calculation of the system matrix. The 142P byte system matrix was compressed into 13.4GB by (1) reducing zero elements, (2) applying the 3D-expanded DOI compression method, (3) factorizing with respect to ring differences and (4) restricting the maximum ring difference to 54 (with only 10% loss of the number of LORs). Histogram-based 3D OSEM based on geometrical system modeling was implemented on a single Itanium 1.6GHz PC with 16Gbyte memory. After evaluating basic imaging performance though phantom experiments, a normal volunteer was scanned (100 min past 104MBq FDG injection, duration 66 min) and the first human brain images were obtained.