M4B2  New Detector Materials/Technologies

Thursday, Nov. 5  10:30-12:30  Pacific Salon 1&2

Session Chair:  Dennis Schaart, Delft University of Technology, Netherlands; Suleman Surti, University of Pennsylvania, United States

Show/Hide All Abstracts

(10:30) M4B2-1, Innovative Silicon Photomultiplier (SiPM) Design for Optimal Coincidence Resolving Time

K. O'Neill, C. Jackson, D. Herbert

SensL Technologies, Cork, Ireland

Optimization of the SiPM diode structure design has previously provided significant performance improvements. This paper will show how an optimized diode design can be further improved via careful design of the packaging and connections. SensL has developed a new TSV (Through Silicon Via) SiPM packaging concept that combines a number of design innovations to provide optimal CRT (Coincidence Resolving Time) with the highest possible PDE. A predictive model of the SensL SiPM is presented, which combines circuit analysis at the diode level, with characterization and signal-to-noise analysis at the device level. By varying the SiPM design choices, the model is used to predict the resulting CRT. Design choices include the microcell density, quench resistance, fast terminal capacitance, substrate contacts to the anode, TSV/wire bond contacts to the cathode, and cathode routing. A number of test devices were fabricated to validate the model, including devices to compare TSV and wirebond connection to anode, substrate versus wirebond connection to cathode, the routing of the anode signal on the silicon, and the microcell density. This paper will present the impact of the various design choices as predicted by the model and validated by testing. For optimized silicon, package and contact design choices, an overall CRT improvement of more than 20% is observed. For packaging alone, the new TSV package delivered a 9% performance improvement in CRT over MLP (Molded Leadframe Package) packaged devices, for equivalent silicon designs. This paper will present further breakdowns of the various contributions to the CRT performance.

(10:45) M4B2-2, Basic Performance of Mg Co-Doped New Scintillator Used for TOF-DOI-PET Systems

T. Kobayashi1,2, S. Yamamoto1, S. Okumura1, J. Y. Yeom3, K. Kamada4

1Radiological and Medical Laboratory Sciences, NAGOYA University, Nagoya, JAPAN
2Department of Radiology, Daiyukai General Hospital, Ichinomiya, JAPAN
3School of Biomedical Engineering, Korea University, Seoul, South Korea
4New Industry Creation Hatchery Center (NICHe), Tohoku University, Sendai, JAPAN

Phoswich depth-of-interaction (DOI) detector using multiple scintillators with different decay time is a useful configuration for developing a high spatial resolution, high sensitivity PET scanner. However, there is not much combination of scintillators suitable for phoswich detectors based on pulse shape discrimination. Recently, a new scintillator with fast decay time, Ce doped Gd3Ga3Al2O12 (GFAG), was developed. It has properties similar to Ce doped Gd3Al2Ga3O12 (GAGG), which is another promising scintillator for PET detector with high light yield. By combing these scintillators, it may be possible to realize a high spatial resolution and fine timing resolution phoswich DOI detector. Moreover, such phoswich DOI detector can be applied to time-of-flight (TOF) systems with good timing performance. Therefore, in this study, we tested the basic performance of the new scintillator-GFAG. GFAGs of 2.9mm x 2.9mm cross-sectional area with 2.9mm, 10mm and 20mm height elements were used for measurements. Although the energy resolution of 9.4% with 10mm height GFAG element was slightly worse than that of GAGG of the same height, the decay time with 59.7ns was faster than that of GAGG with 96.4ns for 662keV gamma photons. Timing resolution measured with a pair of the GFAG with 2.9mm height optically coupled to silicon photomultipliers (Si-PM) was 210 ps FWHM. Timing resolution measured with the 20mm height GFAG/GAGG phoswich pixels was 466 ps. These results indicate that the GFAG is a promising candidate for TOF DOI PET systems.

(11:00) M4B2-3, Pulse Tagging Multiplexing: Single Transmission Line Readout Method for Silicon Photomultiplier TOF-DOI PET

G. B. Ko, J. S. Lee

Department of Nuclear Medicine, Seoul National University, Seoul, Korea

We proposed a novel single transmission line readout method with minimized performance degradation for one-to-one coupled phoswich-type TOF-DOI detector. The basic idea of the proposed multiplexing method, pulse tagging multiplexing (PTM) is adding specially prepared tag signal ahead of the scintillation pulse. The tag signal is a square pulse which encodes photon arrival time and channel information. The two-dimensional position (X and Y) of SiPM array is encoded by the specific width (X) and height (Y) of tag signal. The tag signals and scintillation pulses of each channel are merged by summing amplifier and the final output signal can be read with one-channel fast digitizer. A proof-of-concept detector was designed which consists of 4 × 4 array of LYSO crystals and 16 channel SiPM, and its performance was evaluated. The sixteen 3.12 mm crystals were clearly separated in the flood image with high DWR (DWRx = 14.27 ± 3.84, DWRy = 17.36 ± 4.07). The average energy resolution and CRT were 11.16 ± 1.09% and 395.47 ± 21.5 ps, respectively. A phoswith-type double-layer crystals with two LGSO crystals which had different levels of lutetium contents (t = 40 ns for L0.95GSO and 60 ns for L0.2GSO) to show the feasibility of DOI measurement using pulse shape discrimination. The two type of LGSO crystals are clearly identified with the high reliability (average peak-to-valley ratio was 8.4) thanks to the well preserved shape information in the scintillation pulse with the tagged signal. The proposed multiplexing method allows decoding three dimensional interaction position of gamma-ray in the scintillation detector with single line readout. Energy resolution, CRT, and pulse shape are well preserved showing the promising TOF and DOI capability. We further plan to evaluate TOF phoswich detector consisting of fast scintillation crystals using the PTM method.

(11:15) M4B2-4, High Performance APDs for Time-of-Flight PET

P. Dokhale, M. McClish, R. Farrell, K. Shah

Radiation Monitoring Devices Inc.,, Watertown, MA, USA

Positron Emission Tomography (PET) is a functional imaging technique which can provide diagnosis of diseases such as cancer, Alzheimer’s disease, head trauma, and stroke. Like most imaging modalities, positron emission tomography (PET) is also limited by statistical noise. The statistical noise in clinical PET can be reduced by an order of magnitude by using time-of-flight (TOF) information. This can be achieved by improving the coincidence timing resolution in PET cameras to ~500 ps (FWHM) or better. One of the challenges in developing the TOF PET detector is to find an ideal photodetector with fast response, high quantum efficiency, high energy and timing resolution, compact size, no susceptibility to magnetic fields and low cost. We are exploring new designs of our APD technology and successfully improved the temporal response, particularly at the critical blue end of the spectrum where LSO and LaBr3:Ce emit. We explored the replacement of the relatively thick p-side drift region in our traditional devices with a very thin epitaxial layer having a steep p-type dopant gradient. These new epi-APDs when irradiated with a 405 nm laser showed about 25 times faster rise-time than the traditional APDs. The thinner, p-doped epitaxial layer in epi-APDs (~3 µm) replaces the much thicker drift region (~15 µm) in traditional APDs and provides faster charge collection with reduced charge carrier losses from recombination. Even at this very early stage of investigation, very high coincidence timing resolution of 360 ps (FWHM) was recorded by coupling a new epi-APD (2x2 mm2) to a matching LaBr3:Ce crystal and operating this detector in coincidence with a PMT-LYSO detector. For comparison the same LaBr3 crystal was then coupled to a standard APD of same size and the timing resolution measured was 2.1ns (FWHM). This was an excellent improvement in timing response and we fully expect that upon optimization of these devices and readout electronics, we will be able to improve their coincidence timing resolution to <200 ps (FWHM) for LSO and LaBr3:Ce crystals.

(11:30) M4B2-5, Initial Results for Automatic Calibration of the LabPET II Front-End Detector Module

N. Jürgensen1, L. Arpin1, H. Bouziri1, L. Njejimama1, K. Koua1, E. Gaudin2, J.-F. Pratte1, R. Lecomte2, R. Fontaine1

1Institut Interdisciplinaire d'Innovation Technologique (3IT), Université de Sherbrooke, Sherbrooke, Québec, Canada
2Department of Nuclear Medicine and Radiobiology, Université de Sherbrooke, Sherbrooke, Québec, Canada

An automatic calibration process for the LabPET II detector front-end module has been developed. By aiming at sub-millimetric spatial resolution, an unprecedented channel density is reached. The new detector front-end module is based on an application-specific integrated circuit (ASIC) implementing a Time-over-Threshold (ToT) scheme to extract both energy and time information. Consequently manual channel calibration becomes a tedious task and an automatic calibration strategy adapted to the ToT scheme is required to cope with the huge number and complexity of detector channels. The calibration process involves four main tasks: Clock Phase Adjustment, Photopeak Alignment, Timing Optimization and Energy Correction. To grant an easy and frequent adjustment, the calibration routine must be automatic and real-time. It was implemented in VHDL and C on a Microblaze microprocessor core embedded in a field programmable gate array (FPGA). Misalignment between channels were reduced after calibration from 17% FWHM to 6% FWHM.

(11:45) M4B2-6, Optics Based Method for Ionizing Radiation Photon Detection in PET

L. Tao1, H. Daghighian2, C. S. Levin1,2,3,4

1Dept. of Electrical Engineering, Stanford University, Stanford, CA, USA
2Dept. of Radiology, Stanford University, Stanford, CA, USA
3Dept. of Physics, Stanford University, Stanford, CA, USA
4Dept. of Bioengineering, Stanford University, Stanford, CA, USA

Abstract - In this paper, we prove the feasibility of a totally new optics-based method for ionizing radiation photon detection in PET. Using scintillation detection, the fundamental limit in PET time resolution is strongly dependent on the inherent temporal dispersion of the scintillation mechanism, yielding coincidence time resolution of around 200 ps for 20 mm length crystals. On the other hand, modulation mechanisms of the optical properties of a crystal can be orders of magnitude faster. Therefore we have the goal to borrow from the field of nonlinear optics to study whether ionizing radiation can also produce fast modulation of optical properties. In this paper we demonstrate an optical setup that can detect an ionizing photon induced refractive index change as small as 10-6. Experimental data show that an estimated refractive index change of around 5×10-6 is induced by low flux isotope sources and is successfully detected. Furthermore, we show that the amplitude of the optical modulation signal is linearly proportional to the incoming gamma ray flux with linear fit R factors equal to 0.977 and 0.905, respectively for a Ba-133 and Ge-68 source. Additionally, the normalized signal amplitude is linearly proportional to the average photon energy with a linear fit R factor equal to 0.939. Keywords: optics, optical modulation, ToF PET, time resolution, Cadmium Telluride, electro-optic effect, Pockels effect

(12:00) M4B2-7, Advances in iQID: Upgraded Algorithms, Thicker Scintillators and Larger Area

L. Han1, B. W. Miller1,2, H. B. Barber1,3, L. R. Furenlid1,3

1College of Optical Sciences, University of Arizona, Tucson,AZ, USA
2Pacific Northwest National Laboratory, Richland, WA, USA
3Department of Medical Imaging, University of Arizona, Tucson,AZ, USA

Abstract?iQID (intensified quantum imaging detector) is a novel CCD/CMOS-based ultra-high-resolution photon-counting gamma-ray and x-ray detector technology recently developed at the Center for Gamma-Ray Imaging (CGRI) at the University of Arizona. In this work, we report further advances in iQID?s capabilities in terms of dark-noise suppression, sensitivity and field of view (FOV). A new frame-parsing algorithm has been developed that is capable of eliminating certain kinds of dark noise originating in the image intensifier while at the same time maintaining high processing speed for photon-counting applications. To improve detector sensitivity, a thicker 1.65 mm columnar CsI(Tl) scintillator has been tested and compared against the original 450 ?m columnar CsI(Tl) scintillator in terms of detector sensitivity and resolution. A study of the depth-of-interaction effects of the thicker scintillator has been performed. Simulation and experimental results are in agreement. Finally, to facilitate the use of iQID technology for clinical applications, four large-magnification fiber-optic tapers were tiled together to increase the total detection area to 188?188 mm2. When all of the upgrades are combined, a novel high-sensitivity, high-resolution and large-FOV CCD/CMOS-based gamma-ray detector prototype will be available for evaluation in preclinical and clinical applications.

(12:15) M4B2-8, Development of High-Precision Color Gamma-Ray Image Sensor Based on TSV-MPPC and Diced Scintillator Arrays

T. Oshima1, J. Kataoka1, A. Kishimoto1, T. Fujita1, Y. Kurei1, T. Nishiyama1, S. Yamamoto2, K. Ogawa3

1Science and Engineering, Waseda.Univ., Shinjuku, Tokyo, Japan
2Medicine, Nagoya.Univ., Nagoya-shi, Aichi, Japan
3Science and Engineering, Hosei.Univ., Koganei, Tokyo, Japan

We developed a high-precision color gamma-ray image sensor with fine spatial resolution that is cost-effective, widely applicable, and very sensitive by using a diced cerium-doped Gd3Al2Ga3O12 (Ce:GAGG) scintillator array coupled with a 3.0 × 3.0 pixel 8 × 8 MPPC-array. The proposed image sensor can measure the energy of individual X-ray photons transmitted through an object. The pixel size of the Ce:GAGG scintillator array is 0.2 mm, and the pixels are separated by 50-µm-wide micro-grooves. The image sensor has a size of 20 × 20 mm2 and thickness of 1.0 mm, and it achieved an excellent spatial resolution of 0.3-0.4 mm and energy resolutions of 12% and 18% (FWHM) for 122 and 59.5 keV gamma-rays, respectively. We conducted an experiment to determine the local effective atomic number of metals by using dual-energy gamma-ray sources. In addition, we developed a color-composite image using mixed images taken at three energies (31, 59.5, and 88 keV). Finally, we discuss the application of our novel image sensor for next-generation computerized tomography (CT), i.e., photon-counting CT.