Search

PURPOSE: Previously, we demonstrated the potential of positron emission tomography detectors consisting of monolithic scintillation crystals read out by arrays of solid-state light sensors. We reported detector spatial resolutions of 1.1-1.3 mm full width at half maximum (FWHM) with no degradation for angles of incidence up to 30 degrees, energy resolutions of approximately 11% FWHM, and timing resolutions of approximately 2 ns FWHM, using monolithic LYSO:Ce3+ crystals coupled to avalanche photodiode (APD) arrays. Here we develop, validate, and demonstrate a simple model of the detector point spread function (PSF) of such monolithic scintillator detectors. METHODS: A PSF model was developed that essentially consists of two convolved components, one accounting for the spatial distribution of the energy deposited by annihilation photons within the crystal, and the other for the influences of statistical signal fluctuations and electronic noise. The model was validated through comparison with spatial resolution measurements on a detector consisting of an LYSO:Ce3+ crystal read out by two APD arrays. RESULTS: The model is shown to describe the measured detector spatial response well at the noise levels found in the experiments. In addition, it is demonstrated how the model can be used to correct the measured spatial response for the influence of the finite diameter of the annihilation photon beam used in the experiments, thus obtaining an estimate of the intrinsic detector PSF. CONCLUSIONS: Despite its simplicity, the proposed model is an accurate tool for analyzing the detector PSF of monolithic scintillator detectors and can be used to estimate the intrinsic detector PSF from the measured one.

Item does not contain fulltext Much research is being conducted on position-sensitive scintillation detectors for medical imaging, particularly for emission tomography. Monte Carlo simulations play an essential role in many of these research activities. As the scintillation process, the transport of scintillation photons through the crystal(s), and the conversion of these photons into electronic signals each have a major influence on the detector performance; all of these processes may need to be incorporated in the model to obtain accurate results. In this work the optical and scintillation models of the GEANT4 simulation toolkit are validated by comparing simulations and measurements on monolithic scintillator detectors for high-resolution positron emission tomography (PET). We have furthermore made the GEANT4 optical models available within the user-friendly GATE simulation platform (as of version 3.0). It is shown how the necessary optical input parameters can be determined with sufficient accuracy. The results show that the optical physics models of GATE/GEANT4 enable accurate prediction of the spatial and energy resolution of monolithic scintillator PET detectors.

Neural Networks trained with error back propagation Levenberg-Marquardt training (LM), Neural Networks trained with an algebraic method and Support Vector Machines (SVM) were evaluated to extract the position information from measured light distributions generated by the interactions of 511 keV photons in monolithic scintillator blocks. All three algorithms can achieve a similar average resolution (~1.6 mm FWHM in a 20 times10 times 10 mm LSO block) but the LM trained neural networks do so most efficiently. When the incidence angle of the photons increases to 30deg, the resolution degrades slightly to 2.0 mm FWHM. A small mismatch (< plusmn5deg) between the true incidence angle and the angle for which a neural network was trained can be tolerated without significant resolution loss. Increasing the thickness to 20 mm and using a top-bottom readout of the block yields an average resolution of 2.2 mm FWHM.

Monolithic scintillator blocks provide an efficient way to significantly increase the sensitivity of high-resolution positron emission tomography (PET) systems. Using statistical or neural network-based positioning schemes, the measured intrinsic detector resolution for perpendicular incident photons is 1.7 mm full-width at half-maximum (FWHM) for 10 mm thick LSO blocks and 2.0 mm FWHM for 20 mm thick LSO blocks. Because the positioning algorithms determine the incidence position as opposed to the combination of interaction position and interaction depth, a very good parallax correction can be achieved. Energy and detector time resolution were 11.5% FWHM and 1.6 ns FWHM, respectively.

The absence of very small crystal pixels in monolithic scintillation detectors has a number of potential advantages such as higher sensitivity, better energy resolution and continuous coordinates. In such detectors, the photon incidence position on the detector surface is derived from the measured scintillation light distribution using artificial neural networks (NNs). To this end, each detector module has to be position-calibrated by training the NNs. An automated procedure to simultaneously obtain the calibration data to train NNs for all detector modules in a fully assembled PET system has been developed and evaluated on a simulator set-up.

Previous experimental results have demonstrated that a PET detector module based on a 20times10times10 mm LSO block read out by a Hamamatsu S8550 APD array is able to determine the impinging position of perpendicularly incident 511-keV photons with a resolution better than 2-mm FWHM. A prototype PET demonstrator using two of these detector heads was build and evaluated. The axial and transaxial resolution measured in two-dimensional (2-D) OSEM and 2-D FBP reconstructed images is better than 1.5- and 2-mm FWHM, respectively, for radial distances less than 15 mm. A Monte Carlo simulation showed that a four-ring PET scanner based on these monolithic scintillator blocks has a sensitivity which is about 2.5 times higher than the sensitivity of a similar system equipped with detector heads using a 4times8 matrix of 2times2times10 mm LSO pixels

Item does not contain fulltext Cramér-Rao theory can be used to derive the lower bound on the spatial resolution achievable with position-sensitive scintillation detectors as a function of the detector geometry and the pertinent physical properties of the scintillator, the photosensor and the readout electronics. Knowledge of the Cramér-Rao lower bound (CRLB) can for example be used to optimize the detector design and to test the performance of the method used to derive position information from the detector signals. Here, this approach is demonstrated for monolithic scintillator detectors for positron emission tomography. Two detector geometries are investigated: a 20 × 10 × 10 mm(3) and a 20 × 10 × 20 mm(3) monolithic LYSO:Ce(3+) crystal read out by one or two Hamamatsu S8550SPL avalanche photodiode (APD) arrays, respectively. The results indicate that in these detectors the CRLB is primarily determined by the APD excess noise factor and the number of scintillation photons detected. Furthermore, it is shown that the use of a k-nearest neighbor (k-NN) algorithm for position estimation allows the experimentally obtained spatial resolution to closely approach the CRLB. The approach outlined in this work can in principle be applied to any scintillation detector in which position information is encoded in the distribution of the scintillation light over multiple photosensor elements.

We developed positron emission tomography (PET) detectors based on monolithic scintillation crystals and position-sensitive light sensors. Intrinsic depth-of-interaction (DOI) correction is achieved by deriving the entry points of annihilation photons on the front surface of the crystal from the light sensor signals. Here we characterize the next generation of these detectors, consisting of a 20 mm thick rectangular or trapezoidal LYSO:Ce crystal read out on the front and the back (double-sided readout, DSR) by Hamamatsu S8550SPL avalanche photodiode (APD) arrays optimized for DSR. The full width at half maximum (FWHM) of the detector point-spread function (PSF) obtained with a rectangular crystal at normal incidence equals approximately 1.05 mm at the detector centre, after correction for the approximately 0.9 mm diameter test beam of annihilation photons. Resolution losses of several tenths of a mm occur near the crystal edges. Furthermore, trapezoidal crystals perform almost equally well as rectangular ones, while improving system sensitivity. Due to the highly accurate DOI correction of all detectors, the spatial resolution remains essentially constant for angles of incidence of up to at least 30 degrees . Energy resolutions of approximately 11% FWHM are measured, with a fraction of events of up to 75% in the full-energy peak. The coincidence timing resolution is estimated to be 2.8 ns FWHM. The good spatial, energy and timing resolutions, together with the excellent DOI correction and high detection efficiency of our detectors, are expected to facilitate high and uniform PET system resolution.

Several positioning algorithms are tested to extract position information from the measured scintillation light distribution generated in monolithic LSO blocks of various shapes and read out by a Hamamatsu S8550 APD array. The intrinsic detector resolutions of photons impinging at different angles e.g. 0deg, plusmn 10deg, plusmn 20deg, plusmn 30deg are studied. To this end, we evaluate the following positioning algorithms : Neural Networks trained with error back propagation (Levenberg-Marquardt), Neural Networks trained with an algebraic method and Support Vector Machines (SVM).

The performance of a small-animal positron emission tomography (PET) scanner based on monolithic scintillation detectors read-out by avalanche photo-diode arrays has been investigated by simulation. By minimizing dead space, both within and between the modules, these detectors offer increased detection efficiency compared to pixellated detectors. The spatial resolution of the scanner was investigated in 2-D by simulating a point source at various radial distances from the center. To model the detector response, measured detector line-spread functions were used. An optimum value of approximately 1 mm FWHM was found at 10 mm radial distance from the scanner central axis. Point-source sensitivity profiles in the radial and axial directions were simulated at 1 MBq activity using the Monte-Carlo code GATE. They indicated that monolithic designs increase the sensitivity roughly by a factor of two compared to pixellated designs. NECR curves simulated for these scanner designs show no significant degradation of the performance for activities up to 40 MBq.

The information embedded in the light distribution generated in monolithic scintillator blocks allows the incidence position of an impinging 511 keV gamma to be estimated with an average resolution of 1.8 mm FWHM. To study which instrumentation parameters limit the performance, a validated Monte Carlo simulation was used. It was concluded that the preamplifier noise of 600 electrons rms was the most limiting factor since all other parameters in the set-up such as a photon beam width of 1 mm FWHM, beam step size of 250 mum and an ADC resolution of 12 bits are already at an optimal level. The indiscernible presence of events that undergo multiple interactions in the scintillator block also adds 0.2 mm to the achievable FWHM resolution.

Monolithic scintillation detectors are very promising for high resolution and high sensitivity positron emission tomography. These detectors consist of a few cubic centimeters of scintillating material coupled to one or more position-sensitive APD arrays. The entry point of an impinging annihilation photon is estimated from the light distribution on the APD pixels. In this paper, a model will be derived that predicts the line spread function of these detectors. This model includes the influences of the finite width of the measurement beam, scatter of radiation inside the detector, light yield and intrinsic energy resolution of the scintillator, quantum efficiency, gain and excess noise factor of the avalanche photo-diode arrays, and electronic noise. With this model a better understanding of the behavior of the detector is possible, which can make optimization more efficient.

Recent years have seen an increased interest in high resolution, high sensitivity detectors for small animal positron emission tomography (PET). Detectors based on large (e.g. several cm/sup 3/ of LSO) monolithic scintillators optically coupled to one or more avalanche photodiode (APD) array(s) minimize loss of sensitivity due to dead space. In such detectors, the primary event position and depth of interaction (DOI) correction are derived from the measured distribution of scintillation light over the APD array(s). The performance of detectors consisting of a monolithic LYSO scintillator read out by two Hamamatsu S8550 APD arrays is investigated. Monte Carlo simulations indicate that the scanner sensitivity and uniformity are improved when trapezoidal crystals are used instead of rectangular ones. The performance of both types of detectors was studied. Experiments show that the influence of the angle of incidence on the spatial resolution of both detector types is small, ranging from /spl sim/2.0 mm FWHM at 0/spl deg/ to /spl sim/2.4 mm FWHM at 30/spl deg/. An energy resolution of 11.5% at 511 keV and a time resolution of 2.4 ns was measured. Monte Carlo simulations predict a sensitivity of /spl sim/21% for a point source at the center of the field of view for a four-ring scanner based on trapezoidal monolithic detectors.

Minimizing dead space is one way to increase the detection efficiency of small-animal PET scanners. By using monolithic scintillator crystals (e.g., 20 mm/spl times/10 mm/spl times/10 mm LSO), loss of efficiency due to inter-crystal reflective material is minimized. Readout of such crystals can be performed by means of one or more avalanche photo-diode (APD) arrays optically coupled to the crystal. The entry point of a gamma photon on the crystal surface can be estimated from the measured distribution of the scintillation light over the APD array(s). By estimating the entry point, correction for the depth-of-interaction (DOI) is automatically provided. We are studying the feasibility of such detector modules. To this end, a 64-channel test setup has been developed. Experiments to determine the effect on the spatial resolution of crystal surface finish and detector geometry have been carried out. The first results of these experiments are presented and compared to simulation results. The crystal surface finish has only a small influence on the spatial resolution. The spatial resolution of 20 mm/spl times/10 mm/spl times/10 mm detectors is significantly better when read out on the front side than when read out on the back side. With a 20 mm/spl times/10 mm/spl times/20 mm crystal coupled to two APD arrays, a very small resolution degradation of only /spl sim/0.2 mm is observed for an incidence angle of 30/spl deg/ compared to normal incidence.

Monolithic scintillator detectors, consisting of several cm3 of scintillating material coupled to one or more Hamamatsu S8550 avalanche photodiode (APD) arrays, have been proposed as detectors for high resolution positron emission tomography (PET). These detectors eliminate the dead space caused by reflective material between crystal pixels, and provide good depth of interaction (DOI) information. The entry point of impinging annihilation photons on the front surface of the detector is derived from the scintillation light distribution on the APD array(s). In this work, the various contributions to the energy and timing resolutions were analysed, as well as the influence of several contributions to the spatial resolution. Energy resolutions of around 11.5% FWHM at 511 keV were measured, the major part of which was found to be due to the intrinsic photon variance of the scintillator. A timing resolution of 1.6 ns FWHM was measured against a BaF2-PMT detector. The timing resolution was found to be strongly gain dependent. Spatial resolutions of around 1.75 mm FWHM were measured, fairly independent of the APD gain. This figure still needs to be corrected for the ~0.9 mm FWHM measurement beam.

PET-detectors based on monolithic scintillator blocks have the potential to significantly improve the sensitivity. The impinging photon position of the 511 keV photons on the detector surface is derived from the scintillation light distribution measured by an APD array. To extract this position information, we used Neural Networks (NN). To this end, each detector module has to be position-calibrated by training neural networks. The neural networks can immediately yield a parallax insensitive incidence position of the photon when it was trained with sample data of photons impinging under a similar angle. Different procedures to obtain the calibration data in an automated way on a fully assembled PET system have been developed and validated on a simulator set-up.

We are investigating the possibility of using monolithic scintillator blocks as detectors for small animal positron emission tomography (PET). These detectors consist of several cm/sup 3/ of scintillating material, read out by one or more avalanche photo-diode (APD) arrays. The entry point of an incoming gamma photon is estimated from the distribution of the scintillation light over the APD pixels. To optimize the detector design, the influence of different design parameters is investigated using Geant4 simulations. To make it possible to study the influence of individual design parameters on the intrinsic spatial resolution of the detector, the use of a performance measure is proposed that is independent of the algorithm used to estimate the entry point, namely the Cramer-Rao lower bound on the estimation of the coordinates of a point source of light inside the crystal. To illustrate the use of this method, the influence of optical transport inside the detector is investigated for different detector designs, surface finishes and APD pixel sizes. A comparison with resolutions obtained from simulations involving beams of 511 keV annihilation photons indicates that this approach gives valid results.

Previous experimental results have demonstrated that a PET detector head based on a 20/spl times/10/spl times/10 mm LSO block read out by a Hamamatsu S8550 APD array is able to determine the impinging position of perpendicularly incident 511 keV photons with a resolution better than 2 mm FWHM. A prototype demonstrator using two of these detector heads was build to collect tomographic data. The axial and transaxial resolution measured in 2D OSEM and 2D FBP reconstructed images is better than 1.5 and 2 mm FVVHM respectively for radial distances less than 15 mm. A Monte Carlo simulation showed that a four-ring scanner based on these monolithic scintillator blocks has a 50% higher detection efficiency for single 511 keV photons in comparison to a similar system equipped with detector heads using a 4/spl times/8 matrix of 2/spl times/2/spl times/10 mm LSO pixels.

Monte Carlo simulation is an essential tool in emission tomography that can assist in the design of new medical imaging devices, the optimization of acquisition protocols and the development or assessment of image reconstruction algorithms and correction techniques. GATE, the Geant4 Application for Tomographic Emission, encapsulates the Geant4 libraries to achieve a modular, versatile, scripted simulation toolkit adapted to the field of nuclear medicine. In particular, GATE allows the description of time-dependent phenomena such as source or detector movement, and source decay kinetics. This feature makes it possible to simulate time curves under realistic acquisition conditions and to test dynamic reconstruction algorithms. This paper gives a detailed description of the design and development of GATE by the OpenGATE collaboration, whose continuing objective is to improve, document and validate GATE by simulating commercially available imaging systems for PET and SPECT. Large effort is also invested in the ability and the flexibility to model novel detection systems or systems still under design. A public release of GATE licensed under the GNU Lesser General Public License can be downloaded at http:/www-lphe.epfl.ch/GATE/. Two benchmarks developed for PET and SPECT to test the installation of GATE and to serve as a tutorial for the users are presented. Extensive validation of the GATE simulation platform has been started, comparing simulations and measurements on commercially available acquisition systems. References to those results are listed. The future prospects towards the gridification of GATE and its extension to other domains such as dosimetry are also discussed. ispartof: Physics in medicine and biology vol:49 issue:19 pages:4543-4561 ispartof: location:England status: published

Minimizing dead space is one way to increase the detection efficiency of small-animal PET scanners. By using monolithic blocks of scintillating material (e.g. 20 mmtimes10 mmtimes10 mm LSO), loss of efficiency due to inter-crystal reflective material is minimized. Readout of such blocks can be performed by means of one or more avalanche photo-diode (APD) arrays optically coupled to the block. The primary event position and depth of interaction (DOI) information are derived from the measured distribution of the scintillation light over the APD array(s). We are studying the feasibility of such detector modules, both by simulation and by experiment. A 64-channel setup for testing the above type of detector modules has been developed. Experiments to verify the effect of crystal surface finish, detector geometry and reconstruction algorithm parameters on the spatial resolution have been carried out. The first results of these experiments are presented in this paper, and compared to simulation results. This research is conducted in collaboration with the Crystal Clear Collaboration (CCC)

We are currently investigating the possibility of using continuous scintillator blocks as detectors for small animal PET. These block detectors consist of a few cm 3 of scintillating material, read out by one or two APD arrays. The entry point of an incoming gamma photon is reconstructed from the distribution of the scintillation light over the APD pixels. To optimize the de- tector design, the influence of different parameters is investigated using simulations. As a measure of detector performance that is independent of the reconstruction algorithm used, the Cram´ er- Rao lower bound on the reconstruction of the coordinates of a point source of light inside the crystal is calculated. Results are presented for different detector designs, surface finishes and pixel sizes. A comparison with resolutions obtained from simulations involving beams of 511 keV annihilation photons indicates that this approach gives valid results.