Search

Objective. Monolithic scintillator crystals coupled to silicon photomultiplier (SiPM) arrays are promising detectors for PET applications, offering spatial resolution around 1 mm and depth-of-interaction information. However, their timing resolution has always been inferior to that of pixellated crystals, while the best results on spatial resolution have been obtained with algorithms that cannot operate in real-time in a PET detector. In this study, we explore the capabilities of monolithic crystals with respect to spatial and timing resolution, presenting new algorithms that overcome the mentioned problems. Approach. Our algorithms were tested first using a simulation framework, then on experimentally acquired data. We tested an event timestamping algorithm based on neural networks which was then integrated into a second neural network for simultaneous estimation of the event position and timestamp. Both algorithms are implemented in a low-cost field-programmable gate array that can be integrated in the detector and can process more than 1 million events per second in real-time. Results. Testing the neural network for the simultaneous estimation of the event position and timestamp on experimental data we obtain 0.78 2D FWHM on the (x, y) plane, 1.2 depth-of-interaction FWHM and 156 coincidence time resolution on a 25 mm Ă— 25 mm Ă— 8 mm Ă— LYSO monolith read-out by 64 3 mm Ă— 3 mm Hamamatsu SiPMs. Significance. Our results show that monolithic crystals combined with artificial intelligence can rival pixellated crystals performance for time-of-flight PET applications, while having better spatial resolution and DOI resolution. Thanks to the use of very light neural networks, event characterization can be done on-line directly in the detector, solving the issues of scalability and computational complexity that up to now were preventing the use of monolithic crystals in clinical PET scanners. [ABSTRACT FROM AUTHOR]

Abstract Organ dedicated PET devices provide improved imaging performance when compared to whole body systems. The present study summarizes the test carried out to study a new detector block designed for an organ dedicated PET system. This block includes three novel components namely the scintillator geometry and a retroreflector layer coupled to the entrance face, the photosensor and the readout electronics. We used arrays of 12 × 12 SiPM photosensors with 3 × 3 mm2 active area each and a pitch of 4.2 mm. We are proposing a new readout electronics that permits to reduce the 12 row and columns signals to only 8 without significant detector performance degradation. This approach also allows for resolving radioactive sources in the whole volume of the proposed crystal, significantly reducing the edge effect that typically rejects these events. An overall spatial resolution of about 1.8 mm FWHM is obtained for the whole scintillation volume, with an average energy resolution of 13% FWHM and a photon depth of interaction resolution (FWHM) of 3.7 mm. Highlights • Gamma ray detectors based on thick monolithic scintillators and arrays of SiPMs. • Reduction readout electronics preserving high performance at the crystal edges. • Accurate planar 1.9 mm and DOI 3.7 mm position estimation, and energy resolution 13%. • Detector block suitable for novel PET designs. [ABSTRACT FROM AUTHOR]

In this work, we developed a model able to predict in a few seconds the response of a gamma camera based on continuous scintillator in terms of linearity and spatial resolution in the whole field of view (FoV). This model will be useful during the design phase of a SPECT or PET detector in order to predict and optimize gamma camera performance by varying the parameter values of its components (scintillator, light guides, and photodetector). Starting from a model of the scintillation light distribution on the photodetector sensitive surface, a theoretical analysis based on the estimation theory is carried out in order to find the analytical expressions of bias and FWHM related to four interaction position estimation methods: the classical Center of Gravity method (Anger Logic), an enhanced Center of Gravity method, a Mean Square Error fitting method, and the Maximum Likelihood Estimation method. Afterwards, spatial resolution as well as depth of interaction (DOI) distribution effects are evaluated by processing biases and FWHMs at different DOIs. The comparison between the model and GEANT4 Monte Carlo simulations of four different detection systems has been carried out. Our model prediction errors of spatial resolution, in terms of percentage RMSDs with respect to the simulated spatial resolution, are lower than 13.2% in the whole FoV for three estimation methods. The computational time to calculate spatial resolutions with the model in the whole FoV is five order of magnitudes faster than an equivalent standard Monte Carlo simulation. [ABSTRACT FROM PUBLISHER]

Abstract: In this work we present an innovative algorithm for the reconstruction of PET images based on the List-Mode (LM) technique which improves their spatial resolution compared to results obtained with current MLEM algorithms. This study appears as a part of a large project with the aim of improving diagnosis in early Alzheimer disease stages by means of a newly developed hybrid PET-MR insert. At the present, Alzheimer is the most relevant neurodegenerative disease and the best way to apply an effective treatment is its early diagnosis. The PET device will consist of several monolithic LYSO crystals coupled to SiPM detectors. Monolithic crystals can reduce scanner costs with the advantage to enable implementation of very small virtual pixels in their geometry. This is especially useful for LM reconstruction algorithms, since they do not need a pre-calculated system matrix. We have developed an LM algorithm which has been initially tested with a large aperture (186mm) breast PET system. Such an algorithm instead of using the common lines of response, incorporates a novel calculation of tubes of response. The new approach improves the volumetric spatial resolution about a factor 2 at the border of the field of view when compared with traditionally used MLEM algorithm. Moreover, it has also shown to decrease the image noise, thus increasing the image quality. [Copyright &y& Elsevier]

Improving sensitivity and spatial resolution in small animal Positron Emission Tomography imaging instrumentation constitutes one of the main goals of nuclear imaging research. These parameters are degraded by the presence of gaps between the detectors. The present manuscript experimentally validates our prototype of an edge-less pre-clinical PET system based on a single LYSO:Ce annulus with an inner diameter of 62 mm and 10 outer facets of 26 × 52 mm 2 . Scintillation light is read out by arrays of 8 × 8 SiPMs coupled to the facets, using a projection readout of the rows and columns signals. The readout provides accurate Depth of Interaction (DOI). We have implemented a calibration that mitigates the DOI-dependency of the transaxial and axial impact coordinates, and the energy photopeak gain. An energy resolution of 23.4 ± 1.8% was determined. Average spatial resolution of 1.4 ± 0.2 and 1.3 ± 0.4 mm FWHM were achieved for the radial and axial directions, respectively. We found a peak sensitivity of 3.8% at the system center, and a maximum NECR at 40.6 kcps for 0.27 mCi. The image quality was evaluated using reconstructed images of an array of sources and the NEMA image quality phantom was also studied.

Several groups are actively investigating the performance of monolithic (continuous) scintillation detectors using a variety of crystal thicknesses, photo-sensor configurations, and surface treatments. This work explores the performance of thick LYSO crystals that would be applicable to a whole-body PET system. The crystals were etched with laser induced optical barriers (LIOBs) to alter the behavior of the light spread within the crystal in order to improve the performance of the detector. We studied the behavior of the LIOBs in response to optical light using small cubes of LYSO with a variety of laser etching parameters to characterize the impact of the optical barriers. We demonstrated that the opacity of the etchings can be altered by varying the parameters of the laser etching, which influences the depth-dependent light response and spatial resolution in the thick crystal. We successfully etched several crystals, as large as 50×50×25-mm 3 thick, with a fine grid of LIOBs, and achieved an average spatial resolution close to 3 mm (FWHM) with 511-keV gammas.