Coincidence Time Window Research Articles

Architecture of a dual-modality, high-resolution, fully digital positron emission tomography/computed tomography (PET/CT) scanner for small animal imaging

Contemporary positron emission tomography (PET) scanners are commonly implemented with very large scale integration analog front-end electronics to reduce power consumption, space, noise, and cost. Analog processing yields excellent results in dedicated applications, but offers little flexibility for sophisticated signal processing or for more accurate measurements with newer, fast scintillation crystals. Design goals of the new Sherbrooke PET/computed tomography (CT) scanner are: 1) to achieve 1 mm resolution in both emission (PET) and transmission (CT) imaging using the same detector channels; 2) to be able to count and discriminate individual X-ray photons in CT mode. These requirements can be better met by sampling the analog signal from each individual detector channel as early as possible, using off-the-shelf, 8-b, 100-MHz, high-speed analog-to-digital converters (ADC) and digital processing in field programmable gate arrays (FPGAs). The core of the processing units consists of Xilinx SpartanIIe that can hold up to 16 individual channels. The initial architecture is designed for 1024 channels, but modularity allows extending the system up to 10 K channels or more. This parallel architecture supports count rates in excess of a million hits/s/scintillator in CT mode and up to 100 K events/s/scintillator in PET mode, with a coincidence time window of less than 10 ns full-width at half-maximum.

Ghost interference with an optical parametric amplifier

The 'Ghost' interference experiment is analyzed when the source of entangled photons is a multimode Optical Parametric Amplifier(OPA) whose weak limit is the two-photon Spontaneous Parametric Downconversion(SPDC) beam. The visibility of the double-slit pattern is calculated, taking the finite coincidence time window of the photon counting detectors into account. It is found that the coincidence window and the bandwidth of light reaching the detectors play a crucial role in the loss of visibility on coincidence detection, not only in the 'Ghost' interference experiment but in all experiments involving coincidence detection. The differences between the loss of visibility with two-mode and multimode OPA sources is also discussed. PACS: 42.65.Yj, 42.50.Dv, 42.65.Lm

Simulation of time curves in small animal PET using GATE

The ClearPET project of the Crystal Clear Collaboration (CCC) is building spin-off technology for high resolution small animal Positron Emission Tomography (PET). Monte Carlo simulation is essential for optimizing the specifications of these systems with regards to their most important characteristics, such as spatial resolution, sensitivity, or count rate performance. GATE, the Geant4 Application for Tomographic Emission simulates the passing of time during real acquisitions, allowing to handle dynamic systems such as decaying source distributions or moving detectors. GATE output is analyzed on an event-by-event basis. The time associated with each single event allows to sort coincidences and to model dead-time. This leads to the study of time curves for a prospective small animal PET scanner design. The count rates of true, and random coincidences are discussed together with the corresponding Noise Equivalent Count (NEC) rates as a function of some PET scanner specifications such as detector dead time, or coincidence time window.

Assessment of microPET performance in analyzing the rat brain under different types of anesthesia: comparison between quantitative data obtained with microPET and ex vivo autoradiography

MicroPET (positron emission tomography) has been implemented for use in experiments with small animals. However, the quantification and optimal conditions for scanning are not established yet. The aim of this study was to compare the results obtained by microPET with those by ex vivo autoradiography of rat brain slices, based on the 2-[ 18F]fluoro-2-deoxy- d-glucose (FDG) method, and to establish the optimal conditions for scanning. As an example, we examined glucose metabolism in the rat brain under 6 types of anesthesia and in the conscious state. The scanning conditions for the rat brain were (1) use of a 4-mm-thick leaden jacket, (2) an energy window of 350–650 keV, and (3) a coincidence time window of 6 ns. Under these conditions, the quantitative ROI data from microPET showed a good correlation with the corresponding ROI data from FDG autoradiography in the animal study ( r 2 = 0.81). With our protocol, when anesthesia was started 40 min after the FDG injection, the glucose metabolism was almost the same as that in the conscious rat brain.

Time of flight in PET revisited

PET scanners based on LSO have the potential for significantly better coincidence timing resolution than the 6 ns FWHM typically achieved with BGO. This study analyzes the performance enhancements made possible by improved timing as a function of the coincidence time resolution. If 500 ps FWHM coincidence timing resolution can be achieved in a complete PET camera, the following four benefits can be realized for whole-body FDG imaging: 1) the random event rate can be reduced by using a narrower coincidence timing window, increasing the peak NECR by /spl sim/50%; 2) using time-of-flight (TOF) in the reconstruction algorithm will reduce the noise variance by a factor of 5; 3) emission and transmission data can be acquired simultaneously, reducing the total scan time; and 4) axial blurring can be reduced by using TOF to determine the correct axial plane of origin for each event. While TOF was extensively studied in the 1980s, practical factors limited its effectiveness at that time and little attention has been paid to timing in PET since then. As these potential improvements are substantial and the advent of LSO PET cameras gives us the means to obtain them without other sacrifices, efforts to improve PET timing should resume after their long dormancy.

Development of a GSO detector assembly for a continuous blood sampling system

A new input function monitoring system has been developed using the GSO detector assembly for both PET and SPELT quantitative studies. Due to the rapid rise time of the pulse (about 10nsec), the coincidence time window can be reduced < 1nsec, reducing contribution of randoms associated with the high background activity surrounding the detector. Large light output improved the energy resolution of approximately 11% for 511keV photons, and 20% at 140 keV, as compared with the BGO detector, enabling the use of this system also in SPELT studies. The paired assembly of crystals provided the absolute sensitivity of approximately 7% for PET and 75% for PET tracers. Multiple arrangement of the paired detectors provided possibility of correcting for the transit time of radioactivity through the catheter tube. This study demonstrated that the present system can be of use in both clinical and small animal studies using SPECT and PET tracers.

Isothermal and heated turbulent upflow in a vertical annular channel – Part I. Experimental measurements

The velocity and thermal fields were measured in isothermal and heated turbulent upflow of liquid Refrigerant-113 through a vertical annular channel of inner to outer radius ratio 0.415. A two-component laser Doppler velocimeter was used for the velocity measurements and, simultaneously, a cold-wire for the temperature measurements. The dimensions of the LDV measuring volume and the cold-wire, and their proximity to each other were important considerations. Also crucial to the measurements were the LDV coincidence time window and the temporal response of the cold-wire. Time-mean axial and radial velocities, axial and radial turbulent intensities, the single-point cross-correlation between axial and radial velocity fluctuations (∼axial Reynolds shear stress), and single-point cross-correlations between axial velocity and temperature fluctuations (∼axial turbulent heat flux) and radial velocity and temperature fluctuations (∼radial turbulent heat flux) were measured. Results are reported for Reynolds numbers at channel inlet of 22,800, 31,500, and 46,400 at annulus inner wall heat fluxes of 0, 9000 and 16,000 W m −2. The measured radial turbulent heat flux distribution is compared with that calculated from an approximate form of the thermal energy balance equation in which the measured mean velocity and temperature values are used. Also reported is the radial distribution of turbulent Prandtl number estimated from the measurements.

The ECAT ART Scanner for Positron Emission Tomography: 1. Improvements in Performance Characteristics

The widespread use of positron emission tomography (PET) has been to some extent limited by the cost and complexity of PET instrumentation. Recognition of the wider applicability of clinical PET imaging is reflected in the ECAT ART design, a low cost PET scanner targeted for clinical applications, particularly in oncology. The ART comprises two asymmetrically opposed arrays of BGO block detectors. Each array consists of 88 (transaxial) by 24 (axial) crystals, and the arrays rotate continuously at 30 rpm to acquire a full 3D projection data set. Sensitivity and count rate limitations are key performance parameters for any imaging device. This paper reports on improved performance characteristics of the ART, achieved by operating the scanner with a decreased block integration time, reduced coincidence time window, and collimated singles transmission sources. Compared to the standard ART configuration, these modifications result in both improved count rate performance and higher quality transmission scans.

Investigation of noise equivalent count rate in positron imaging using a dual head gamma camera

In positron imaging, image quality depends on scatter, random and coincidence rate. It is known that noise equivalent count (NEC) rate is a good indicator of image quality and that it helps optimizing acquisition parameters. We measured the NEC curve with a SophyCamera DST (SMV) using a 20 cm cylinder phantom filled with /sup 18/F (59 MBq). The field of view of the DST is determined by two rectangular parallel detectors (400*300 mm) separated by 720 mm and each scintillation detector is composed of a 3/8 of an inch thick NaI:Tl crystal. The coincidence time window was set to 15 ns and the scatter fraction (sf) was measured using a PET NEMA protocol. We measured random and coincident count rates using tomographic acquisition with two 20% energy windows centered on 511 KeV (photopeak) and 417 KeV (Compton) over 6 half-lives of /sup 18/F. We investigated the random coincidence rate by delaying signals from one of the coincident detectors (45 ns) and measured system sensitivity as the slope at the origin of the NEC curve. The NEC rate peaked at 199 cps for photopeak (with a 0.20 sf) and at 252 cps for photopeak+Compton (with 0.28 sf) coincidences. These maxima were attained over activity concentration values ranging from 1.33 kBq/cc to 1.85 kBq/cc. Our results show that image quality improves when using photopeak+Compton coincidences with a 30% greater sensitivity and an increased NEC rate.

A comparison of the efficiency and accuracy of medial superior olive cell simulations

Binaural processing of interaural time differences (ITDs) is widely believed to be performed by a cross‐correlation mechanism. Cells have been found in the medial‐superior olive (MSO) and inferior colliculus which respond most strongly when presented with stimuli with a characteristic ITD (EE cells). These cells are believed to form the basis for such a mechanism. In this study we compare the performance of cells based on simple cross‐correlation, coincidence detector cells based on a novel deterministic method, and coincidence detectors based on Monte‐Carlo simulation. The deterministic model is based upon the requirement that the probability of a given number of inputs from the auditory nerves of either ear must exceed a threshold amount within a finite time (coincidence window) before the modeled cell can fire. We examine the firing rate and vector strength as a function of ITD, overall level, and interaural level difference (ILD) and compare our results with the data from cells in the MSO. We find that a simple cross‐correlation cell is inadequate to describe the data. However, the deterministic cell and Monte‐Carlo simulation, both with a coincidence window width of 0.2 ms, describe the data very well. [Work supported by The Wellcome Trust, UK.]

Elastic recoil detection analysis using ion-induced electron emission for particle identification

We propose a new method to identify particles in ERD analysis, using their electron emission yield from a thin carbon foil. Before the particles reach a silicon surface barrier detector (SB) they penetrate a set of thin foils (typically 6 foils) with a thickness of 3 μgcm2 each). The emission yield depends on the nuclear charge of the penetrating ion and it is roughly proportional to the energy loss in the foil. The emitted electrons are accelerated to a muchannel plate (MCP) by a voltage of 300 V. The electron signal from the MCP is proportional to the number of emitted electrons and it occurs in coincidence with the energy signal from the energy detector. For data acquisition we developed a dual parameter multichannel analyzer (M2D) as an add on board for an industry standard personal computer. The two-dimensional spectrum of coincidences and the one-dimensional spectra from both detectors are recorded simultaneously. The M2D has 256K channels which can be freely configured as a two-dimensional matrix. For example a resolution of 1024 × 256 channels is possible. For optimum suppression of random coincidences the coincidence time window can be set from 0.125 μs up to 32 μs. For this new setup the ability for particle identification is discussed for different projectiles (He, C, O, Cl) and targets. H recoil ions can be well separated from He projectiles so that for H analysis the H recoil spectrum and the He forward energy spectrum can be measured simultaneously. An example for depth-profiling of 100 keV H implantations in silicon is given.

Digital coincidence detection: a scanning VLSI implementation

The authors have implemented a modular digital coincidence detection circuit in VLSI, which drastically reduces the size and increases the performance of the coincidence detection logic required in a PET (positron emission tomography) scanner. Important acquisition features contained within this integrated solution include: (1) prompt and delayed processing channels for each event-pair, with programmable coincidence time windows; (2) event time difference mode, which allows for a fast and accurate system timing calibration algorithm; (3) programmable axial event acceptance, which provides acquisition flexibility from conventional 2-D through fully 3-D sinogram sets; and (4) programmable transaxial field of view, which provides an electronic collimator to ignore event-pairs outside of the desired field of view. The modularity of the design allows it to be used on various system geometries. >

Circular polarization of the 3D state of atomic hydrogen excited by electrons

The theory of the circular polarization from the 32D state is analysed and angular correlation formulae are derived for the scattered n = 3 electron and the cascade Lyman-alpha radiation. The formulae include allowance for the polarization of the intermediate-state 2p electron from the LS coupling in the time evolution of the decay of the collisionally excited n = 32D states. The effect of a finite coincidence time window is also examined. The predictions of the model are compared with the model which assumes that the intermediate-state 2p electron is unpolarized. New experimental data at an incident electron kinetic energy of 54.4 eV are reported and indicate the circular polarization of the Lyman-alpha radiation has small negative values from 10 degrees to 120 degrees . The theoretical predictions of the correlation are illustrated in a six-state close-coupling model which gives a minimum at forward angles but does not agree quantitatively with the experimental data.

Einstein-Podolsky-Rosen state for space-time variables in a two-photon interference experiment.

A pair of correlated photons generated from parametric down-conversion was sent to two independent Michelson interferometers. Second-order interference was studied by means of a coincidence measurement between the outputs of two interferometers. The reported experiment and analysis studied this second-order interference phenomenon from the point of view of the Einstein-Podolsky-Rosen (EPR) paradox. The experiment was done in two steps. The first step of the experiment used 50-psec and 3-nsec coincidence time windows simultaneously. The 50-psec window was able to distinguish a 1.5-cm optical path difference in the interferometers. The interference visibility was measured to be 38% and 21% for the 50-psec time window and 22% and 7% for the 3-nsec time window, when the optical path differences of the interferometers were 2 cm and 4 cm, respectively. By comparing the visibilities between these two windows, the experiment showed the nonclassical effect which resulted from an EPR state. The second step of the experiment used a 20-psec coincidence time window, which was able to distinguish a 6-mm optical path difference in the interferometers. The interference visibilities were measured to be 59% for an optical path difference of 7 mm. This is an observation of visibility greater than 50% for a two-interferometer EPR experiment which demonstrates nonclassical correlation of space-time variables.

Models of a two-photon Einstein-Podolsky-Rosen interference experiment.

Models of a two-photon interference experiment proposed by Franson to study the Einstein-Podolsky-Rosen (EPR) effect are presented. The experiment considers a pair of photons correlated in time and in frequency, each of which is passed through an interferometer. When the path difference through the two interferometers is set so there is no first-order interference, Franson showed that second-order interference between the two photons could occur with 100% visibility. In this paper both quantum and classical models are considered and the importance of the coincidence time window and the effect of the frequency correlation are shown. Calculations are presented for all settings of the interferometers. The EPR effect arises from entangled states; a brief discussion is given of the nature of the entangled state for this type of experiment.

Hidden variables with directionalization

A hidden-variables model is presented which, by using a hypothesis of “directionalization” of the photons at the deflectors, is able to reproduce all the quantum mechanical predictions for the Orsay realization of the Einstein-Podolsky-Rosen-Bohm experiment, even for ideal polarizers, detectors, time-coincidence windows, and “event-ready” setups. The model also holds for the no-enhancement assumption. The requirements for an experiment aimed to discriminate between quantum mechanics and the new model are discussed. Under some plausible assumptions, such experiment is achievable with the current technology. It would be essentially a remake of the Orsay one with improved deflectors and a time-resolved measurement of the photon flows toward each detector.

Geometric bias and time coincidence in 3-dimensional laser Doppler velocimeter systems

The measurement by a 3-d laser Doppler velocimeter of a turbulent flow has been numerically simulated. Errors associated with the probe volume geometry and the coincidence time window concept are revealed. One type of error occurs for high system data rates when multiple particles lead to system realizations. Another error occurs associated with a geometric bias discovered in the present study. This 3-d Idv geometric bias exists even for single-particle realizations and regardless of the system data rate. A technique for the elimination of the geometric bias is presented.

Position sensitive detectors in (e,2e) coincidence measurements

The two outgoing electrons from (e,2e) scattering processes are selected by two 180° electrostatic analyzers whose energy dispersed output spectra are detected by microchannel plates followed by resistive anodes. These detectors permit the identification of all multiple coincident events occurring within the dispersal width of the analyzers. It was found that the time difference between the two detectors for correlated events is strongly dependent on the energy of the incident electrons, which consequently causes the time coincidence window to move systematically. This behavior has been accounted for by a model that determines the time taken by an electron to go around a hemispherical analyzer. A description is given of a data collection system that can compensate for the shift in the coincidence spectra and that records all multiple coincidence events with an increase in the overall coincidence count rate of about two orders of magnitude relative to conventional non position sensitive detectors without position sensing. The performance of the system is shown for an (e,2e) spectrum of argon.

Initial Results from the Donner 600 Crystal Positron Tomograph

We describe a positron tomograph using a single ring of 600 close-packed 3 mm wide bismuth germanate (BGO) crystals coupled to 14 mm phototubes. The phototube preamplifier circuit derives a timing pulse from the first photoelectron, and sends it to address and coincidence circuits only if the integrated pulse height is within a pre-set window. The timing delays and pulse height windows for all 600 detectors and the coincidence timing windows are computer adjustable. An orbiting positron source is used for transmission measurements and a look-up table is used to reject scattered and random coincidences that do not pass through the source. Data can be acquired using a stationary mode for 1.57 mm lateral sampling or the two-position clam sampling mode for 0.79 mm lateral sampling. High maximum data rates are provided by 45 parallel coincidence circuits and 4 parallel histogram memory units. With two-position sampling and 1.57 mm bins, the reconstructed point spread function (PSF) of a 0.35 mm diam 22Na wire source at the center of the tomograph is circular with 2.9 mm full-width at half-maximum (fwhm) and the PSF at a distance of 8 cm from the center is elliptical with a radial fwhm of 4.0 mm and tangential fwhm of 3.0 mm.

The Use of an Optimized Coincidence Time Window in BGO Positron Cameras

In a positron emission tomograph the ratio of true-to-random coincidences decreases at elevated count rates, thereby causing a degraded image. Efficiency simulations of a representative sample configuration show that for brain cameras, the high sensitivity range of BGO systems can be extended by using an optimized coincidence time window. The window width is then chosen to optimize the signal-to-noise ratio at each count rate level. To further evaluate the improvement we have also compared the BGO system to a BaF2 system.