Sun Il Kwon

An Investigation Into the Use of Geiger-Mode Solid-State Photomultipliers for Simultaneous PET and MRI Acquisition

Photon detecting Geiger-mode solid-state devices are being actively researched and developed because, unlike photo- multiplier tubes (PMT), they can be used in high-magnetic-field and radio-frequency environments, such as in magnetic resonance imaging (MRI) scanners. In addition, some Geiger-mode solid-state devices have higher photon detection efficiencies than PMTs and higher gains than avalanche photo-diodes (APD). We tested Geiger-mode solid-state photomultipliers (SSPM) inside a 3 T MRI to study the possibility of using them in combined PET/MRI scanners. Approximately 16% energy resolutions and ~1.3 ns coincidence time resolutions with 22Na and lutetium yttrium oxyorthosilicate (LYSO) were obtained for full-width at half maximum (FWHM) for T1, T2, and gradient echo T2* MRI pulse sequences with little MR image degradation. The study shows that SSPMs have excellent potential for use in combined PET/MRI scanners.

A Four-Layer DOI Detector With a Relative Offset for Use in an Animal PET System

For animal PET systems to achieve high sensitivity without adversely affecting spatial resolution, they must have the ability to measure depth-of-interaction (DOI). In this paper, we propose a novel four-layer PET system, and present the performances of modules built to verify the concept of the system. Each layer in the four-layer PET system has a relative offset of half a crystal pitch from other layers. Performances of the four-layer detector were estimated using a GATE Monte Carlo simulation code. The proposed system consists of six H9500 PMTs, each of which contains 3193 crystals. A sensitivity of 11.8% was obtained at the FOV center position of the proposed system. To verify the concept, we tested a PET module constructed using a H9500 flat panel PMT and LYSO crystals of cross-sectional area 1.5 × 1.5 mm2. The PET module was irradiated with a 1.8 MBq 22Na radiation source from the front or side of the crystals to obtain flood images of each crystal. Collimation for side irradiation was achieved using a pair of lead blocks of dimension 50 × 100 × 200 mm3. All crystals in the four layers were clearly identified in flood images, thus verifying the DOI capability of the proposed four-layer PET system. We also investigated the optimal combination of crystal lengths in the four-layer PET system using the GATE Monte Carlo simulation code to generate events from simulated radiation sources, and using the ML-EM algorithm to reconstruct simulated radiation sources. The combination of short crystal lengths near radiation sources and long crystal lengths near the PMT provides better spatial resolution than combinations of same crystal lengths in the four-layer PET system.

Concept Verification of Three-Layer DOI Detectors for Small Animal PET

Improved spatial resolution without sacrificing sensitivity is one of the most challenging developmental goals for small animal PET scanners. The 3-layer configuration that we propose here utilizes relative offsets of half a crystal pitch in x- and y-directions, and pulse shape discrimination to obtain depth of interaction (DOI). Three layers of crystals with a dimension 1.5 x 1.5 x 7.0 mm3 were composed of a L0.2 GSO (Lu0.4 Gd1.6 SiO4: Ce) crystal layer and a L0.9 GSO (Lu1.8 Gd0.2 SiO4 Ce) crystal layer aligned with each other, and a L0.9 GSO crystal layer offset at half a crystal pitch in x- and y-directions. The L0.9 GSO crystal layer was attached to a Hamamatsu H9500 flat-panel PMT. The devised small animal PET scanner has a diameter of 84 mm with one detector ring, and can be upgraded to two detector rings. GEANT4 Monte-Carlo simulation was used to estimate sensitivities of ~12% and ~20%, respectively, at the center of one and two PMT ring system with an energy window of 350~750 keV. We present flood images with peak-to-valley ratios of about 5-6 obtained using 22Na and layer identification capability of ~99 % with pulse shape analysis, and verified the basic concepts of multi-layer small animal PET.

Dual-Phase Tapped-Delay-Line Time-to-Digital Converter With On-the-Fly Calibration Implemented in 40 nm FPGA

This paper describes two novel time-to-digital converter (TDC) architectures. The first is a dual-phase tapped-delay-line (TDL) TDC architecture that allows us to minimize the clock skew problem that causes the highly nonlinear characteristics of the TDC. The second is a pipelined on-the-fly calibration architecture that continuously compensates the nonlinearity and calibrates the fine times using the most up-to-date bin widths without additional dead time. The two architectures were combined and implemented in a single Virtex-6 device (ML605, Xilinx) for time interval measurement. The standard uncertainty for the time intervals from 0 to 20 ns was less than 12.83 ps-RMS (root mean square). The resolution (i.e., the least significant bit, LSB) of the TDC was approximately 10 ps at room temperature. The differential nonlinearity (DNL) values were [-1.0, 1.91] and [-1.0, 1.88] LSB and the integral nonlinearity (INL) values were [-2.20, 2.60] and [-1.63, 3.93] LSB for the two different TDLs that constitute one TDC channel. During temperature drift from 10 to 50°C, the TDC with on-the-fly calibration maintained the standard uncertainty of 11.03 ps-RMS.

A novel compensation method for the anode gain non-uniformity of multi-anode photomultiplier tubes

The position-sensitive multi-anode photomultiplier tube (MA-PMT) is widely used in high-resolution scintillation detectors. However, the anode gain nonuniformity of this device is a limiting factor that degrades the intrinsic performance of the detector module. The aim of this work was to develop a gain compensation method for the MA-PMT and evaluate the resulting enhancement in the performance of the detector. The method employs a circuit that is composed only of resistors and is placed between the MA-PMT and a resistive charge division network (RCN) used for position encoding. The goal of the circuit is to divide the output current from each anode, so the same current flows into the RCN regardless of the anode gain. The current division is controlled by the combination of a fixed-value series resistor with an output impedance that is much larger than the input impedance of the RCN, and a parallel resistor, which detours part of the current to ground. PSpice simulations of the compensation circuit and the RCN were performed to determine optimal values for the compensation resistors when used with Hamamatsu H8500 MAPMTs. The intrinsic characteristics of a detector module consisting of this MA-PMT and a lutetium-gadolinium-oxyorthosilicate (LGSO) crystal array were tested with and without the gain compensation method. In simulation, the average coefficient of variation and max/min ratio decreased from 15.7% to 2.7% and 2.0 to 1.2, respectively. In the flood map of the LGSO-H8500 detector, the uniformity of the photopeak position for individual crystals and the energy resolution were much improved. The feasibility of the method was shown by applying it to an octagonal prototype positron emission tomography scanner.

A Feasibility Study on the Use of Optical Fibers for the Transfer of Scintillation Light to Silicon Photomultipliers

Integrated PET/MRI units with simultaneous acquisition capability are set to play an important role in studies of human breast and prostate imaging and brain function. However, to take advantage of existing MRI units in hospitals and institutions, minimally modified combined PET MRI is highly desirable. In addition, the current MRI trend is to utilize powerful body coils to transmit radio-frequency (RF) waves and local RF coils to receive signals. The authors propose a silicon photomultiplier (SiPM) PET equipped with optical fiber bundles that transfer photons from scintillation crystal to SiPM. To investigate the feasibility of SiPM PET using optical fiber bundles, the authors studied the performances of SiPM/scintillator couplings using single optical fibers and a fiber bundle. GEANT4 Monte-Carlo simulation was used to study scintillation photon transfer from scintillation crystals to the SiPM. This simulation showed that light loss, due to the bending of an optical fiber, is not significant for a fiber with a diameter of 2.0 mm and a bending radius of greater than 25 mm. To validate the GEANT4 Monte-Carlo simulation, several simple detectors were assembled and tested. Simulation results agreed reasonably well with experimental results. Two Hamamatsu multi-pixel photon counters (MPPCs) were tested using double clad optical fibers of 1.5 mm and 2.0 mm diameter, and 25 mm and 50 mm bending radius, respectively. When two MPPCs were directly attached to 2.0 × 2.0 × 10.0 mm3 LYSO crystals, a ~14% energy and a ~1.3 ns coincidence timing resolution were obtained at full width half maximum (FWHM). With one of the MPPCs attached to an optical fiber of 1.5 mm diameter, 50 mm bending radius, and length 300 mm, energy and coincidence timing resolutions were 27% and 2.2 ns, respectively. With an optical fiber bundle made of bare fibers with 1.5 mm diameter and length 100 mm, an ~ 26% energy resolution was obtained. Even though the Monte-Carlo simulation showed light loss was not significant for a single 90° turn of bending, the mechanical integrity of the optical fiber, especially the absence of cracks which can be caused by sharp bending, seemed to be a far more important constraint on sharp bending. These initial results are encouraging with respect to the use of combined SiPM PET using optical fibers.

Four-layer DOI detector with a relative offset in animal PET system

In order to build an animal PET system with high sensitivity and resolution, we propose a four-layer PET system which provides the DOI (depth of interaction) information. Each crystal layer has a relative offset of half the crystal pitch with each other and uses an identical type of crystals simplifying readout electronics. The proposed system PET system consists of four LYSO (Lu₁₈Y_0.2SiO₅:Ce) or L_0.9GSO (Lu_1.8Gd_0.2 Si_0.4:Ce) crystal layers with a crystal dimension of 1.5times1.5times5.0 mm³. The animal PET system has a diameter of 84 mm with one Hamamatsu H9500 flat panel PMT ring, and can be upgraded to two PMT ring configuration. An estimated efficiency using GATE simulation is ~15% at the center of one detector ring system. We present flood images obtained with a ²²Na radiation source to verify the concepts of the proposed system, and obtained promising results.

SPATIAL AND ENERGY RESOLUTIONS OF A HEXAGONAL ANIMAL PET SCANNER BASED ON LGSO CRYSTAL AND FLAT-PANEL PMT

The aim of this study was to explore the spatial and energy resolutions of a PET scanner that we have recently developed. The scanner, which consists of six detector modules with 1-layer LGSO crystals, has a hexagonal configuration with a faceto-face distance of 86.4 mm between two opposite PET modules; such properties facilitate the imaging of small animals. A 22Na point source was employed to estimate horizontal and vertical spatial resolutions. To assess the energy resolution, a uniform 18F cylindrical phantom was scanned. A software-based spectrum analysis of list-mode data was used to assign a local energy window centered on the photopeak position for every single crystal. For the image reconstruction, an ML-EM algorithm was used. The spatial resolutions at the center of the scanner were 0.99 mm in the horizontal direction and 1.13 mm in the vertical direction. The energy resolution averaged over each PMT ranged from 13.3%-14.3%, which gave an average value of 13.8%. These results show that this simple system is promising for small animal imaging with excellent spatial and energy resolutions.