Jorge Miguel Tavares Couceiro de Sousa
@isec.pt
Departament of Physics and Mathematics of the Coimbra Institute of Engineering (ISEC - Instituto Superior de Engenharia de Coimbra)
Polythecnic of Coimbra (IPC - Instituto Politécnico de Coimbra)
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Paulo Fonte, Luís Lopes, Rui Alves, Nuno Carolino, Paulo Crespo, Miguel Couceiro, Orlando Cunha, Nuno Dias, Nuno C. Ferreira, Susete Fetal,et al.
Elsevier BV
Ana Luisa Lopes, Miguel Couceiro, Paulo Crespo, and Paulo Fonte
IEEE
Based on previous results obtained with a pre-clinical Positron Emission Tomography (PET) scanner for mice using Resistive Plate Chambers (RPCs) detectors, we believe that constructing a brain-dedicated PET based on RPCs, the HiRezBrainPET, may provide the very-high spatial resolution deemed helpful in brain imaging. Previous experimental studies with this small-animal RPC-PET system provided a spatial resolution of 0.4 mm Full Width at Half Maximum. The brain-dedicated scanner, very similar in shape to the previously proposed for Time-Of-Flight (TOF) Human full-body PET, will consist of 4 detection heads perfectly aligned, each having a stack of a given number of RPC detectors (4, 8, or 10 in the present work). Each detector will consist of 2 detection modules with five amplification gaps, each with its axial electrode, sharing a common transaxial and timing electrode. This work is part of the active development of a first fully functional prototype. It presents the initial studies performed by Monte Carlo simulations to find: 1) the thickness of the RPCs glass plates that maximize the detection efficiency, as well as the influence of the readout electrodes thickness on the detection efficiency; and 2) the loss of coincidence data by only performing coincidences between opposing heads, relative to that obtained if the coincidences were performed using all available heads. The results showed that for stacks of 4, 8, and 10 detectors, the glass plate thickness that maximizes the detection efficiency is equal to, respectively, ~308 µm, ~279 µm, and ~257 µm, with the corresponding detection efficiencies being 7.18%, 12.97%, and 15.47%. For the glass plates thicknesses commercially available – 280 µm, 330 µm, and 400 µm - the efficiencies obtained were, respectively: 7.10%, 7.15%, and 7.09% for a stack of 4 detectors, 12.97%, 12.80%, and 12.40% for a stack of 8 detectors, and 15.40%, 15.12% and 14.59% for a stack of 10 detectors. It was also found that the influence of the readout electrodes on the detection efficiency is negligible. Regarding the coincidences, almost all coincidences occur between opposing heads for sources placed at the scanner center. However, as the source is moved off-center in the transaxial plane, the number of coincidences between opposing heads decreases significantly, increasing the number of coincidences between adjacent heads. Thus, it is worth implementing a coincidence scheme that uses all detection heads instead of only opposing ones.
M. Couceiro, P. Crespo, A. Blanco, N.C. Ferreira, L. Mendes, R. Ferreira Marques, and P. Fonte
Institute of Physics, Polish Academy of Sciences
An Unlikely but Promising Approach M. Couceiro, P. Crespo, A. Blanco, N.C. Ferreira, L. Mendes, R. Ferreira Marques and P. Fonte Laboratorio de Instrumentacao e Fisica Experimental de Particulas, Departamento de Fisica, Universidade de Coimbra, Rua Larga, 3004-516 Coimbra, Portugal Instituto Politecnico de Coimbra, ISEC, Rua Pedro Nunes, Quinta da Nora, 3030-199 Coimbra, Portugal Departamento de Fisica, Faculdade de Ciencias e Tecnologia, Universidade de Coimbra, Rua Larga, 3004-516 Coimbra, Portugal Instituto Biomedico de Investigacao da Luz e Imagem, Faculdade de Medicina, Universidade de Coimbra, Azinhaga Santa Comba, Celas, 3004-516 Coimbra, Portugal Instituto de Ciencias Nucleares Aplicadas a Saude, Faculdade de Medicina, Universidade de Coimbra, Azinhaga Santa Comba, Celas, 3004-516 Coimbra, Portugal
Miguel Couceiro, Paulo Crespo, Rui F. Marques, and Paulo Fonte
Institute of Electrical and Electronics Engineers (IEEE)
Scatter Fraction (SF) and Noise Equivalent Count Rate (NECR) of a 2400 mm wide axial field-of-view Positron Emission Tomography (PET) system based on Resistive Plate Chamber (RPC) detectors with 300 ps Time Of Flight (TOF) resolution were studied by simulation using Geant4. The study followed the NEMA NU2-2001 standards, using the standard 700 mm long phantom and an axially extended one with 1800 mm, modeling the foreseeable use of this PET system. Data was processed based on the actual RPC readout, which requires a 0.2 μs non-paralyzable dead time for timing signals and a paralyzable dead time (τps) for position signals. For NECR, the best coincidence trigger consisted of a multiple time window coincidence sorter retaining single coincidence pairs (involving only two photons) and all possible coincidence pairs obtained from Multiple coincidences, keeping only those for which the direct TOF-reconstructed point falls inside a tight region surrounding the phantom. For the 700 mm phantom, the SF was 51.8% and, with τps = 3.0 μs, the peak NECR was 167 kcps at 7.6 kBq/cm3. Using τps = 1.0 μs the NECR was 349 kcps at 7.6 kBq/cm3, and no peak was found. For the 1800 mm phantom, the SF was slightly higher, and the NECR curves were identical to those obtained with the standard phantom, but shifted to lower activity concentrations. Although the higher SF, the values obtained for NECR allow concluding that the proposed scanner is expected to outperform current commercial PET systems.
Paulo Crespo, Alberto Blanco, Miguel Couceiro, Nuno C. Ferreira, Luís Lopes, Paulo Martins, Rui Ferreira Marques, and Paulo Fonte
Springer Science and Business Media LLC
Resistive plate chambers (RPC) were originally deployed for high energy physics. Realizing how their properties match the needs of nuclear medicine, a LIP team proposed applying RPCs to both preclinical and clinical positron emission tomography (RPC-PET). We show a large-area RPC-PET simulated scanner covering an axial length of 2.4m —slightly superior to the height of the human body— allowing for whole-body, single-bed RPC-PET acquisitions. Simulations following NEMA (National Electrical Manufacturers Association, USA) protocols yield a system sensitivity at least one order of magnitude larger than present-day, commercial PET systems. Reconstruction of whole-body simulated data is feasible by using a dedicated, direct time-of-flight-based algorithm implemented onto an ordered subsets estimation maximization parallelized strategy. Whole-body RPC-PET patient images following the injection of only 2mCi of 18-fluorodesoxyglucose (FDG) are expected to be ready 7 minutes after the 6 minutes necessary for data acquisition. This compares to the 10-20mCi FDG presently injected for a PET scan, and to the uncomfortable 20-30minutes necessary for its data acquisition. In the preclinical field, two fully instrumented detector heads have been assembled aiming at a four-head-based, small-animal RPC-PET system. Images of a disk-shaped and a needle-like 22Na source show unprecedented sub-millimeter spatial resolution.
Paulo Martins, Miguel Couceiro, Nuno C. Ferreira, Rui Ferreira Marques, Paulo Fonte, Luis Mendes, and Paulo Crespo
IEEE
A single-bed, whole-body positron emission tomograph based on resistive plate chamber detectors has been proposed (RPC-PET). It has been shown by simulation that RPCPET with an axial field-of-view (AFOV) of 204m is feasible and yields an absolute sensitivity enhancement of at least one order of magnitude superior to that of typical cylindrical, crystal-based PET scanners. In addition to its time-of-flight (TOF) advantage, RPC-PET offers potential very-high spatial resolution at the detector level. A fully-3D reconstruction algorithm capable of processing the very inclined Iines-of-response (LOR) from large AFOV systems such as RPC-PET has been demonstrated. It relies on the application of a TOF-based-kernel into the maximum likelihood estimation maximization algorithm. By means of a 300 ps full width at half maximum (FWHM) time resolution, a rejection of 77% of the scattered events was obtained. It is shown that the scatter fraction rejection grows exponentially with an increasing time resolution. We present reconstructed results from blind simulations corresponding to an anthropomorphic phantom with oncological lesions of several sizes immersed into different locations within the human body. A comparison between 300 and 600 ps FWHM TOF reconstructed images is performed. An increasing detectability is observed for a better TOF resolution. We finally compare issues related to image convergence speed and computational burden, making use of graphical processing units (GPUs) and 16 threads central processing units (CPUs). GPUs perform better by a factor two in speed. An alternative approach, which consists in dividing the acquired data into five different image regions, which are reconstructed independently, provides a three times faster reconstruction, as compared with whole-body reconstruction, and allowing to reach a reconstructed image by means of a 300ps FWHM RPC-PET scanner in 7 minutes after the end of data acquisition.
Miguel Couceiro, Paulo Crespo, Rui Ferreira Marques, and Paulo Fonte
IEEE
Scatter Fraction (SF) and Noise Equivalent Count Rate (NECR) of a 240 cm wide axial field-of-view Positron Emission Tomography (PET) system based on Resistive Plate Chamber (RPC) detectors with 300 ps Time of Flight (TOF) resolution were studied by detailed simulations using GEANT4 (version 9.2, patch 04), following the NEMA NU2-2001 protocol, both with the standard 70 cm length phantom and an axially extended one (180 cm length). Simulation-produced data were processed to account for detector readout with a τis = 0.2 μs non-paralyzable dead time for time signals followed by a τps paralyzable dead time for position signals, for which pileup events can be rejected or accepted with a coarse position (1 cm σ Gaussian distribution and 3 cm bins in the axial and transaxial directions, respectively). Concerning NECR, the best coincidence trigger strategy consisted on performing a multiple time window coincidence sorter retaining both single coincidence pairs and all single pairs out of multiple coincidences, followed by acceptance of coincidences for which a direct TOF-reconstructed point falls inside a tight region surrounding the phantom. In these conditions, SF was 51.8% for the standard NEMA NU2-2001 phantom and 53.7% for the extended one, and independent of τps. With a value τps = 3.0 μs, and rejecting pileup position events, peak NECR was ~167 kcps at ~7.6 kBq/cm3 for standard NEMA NU2-2001 phantom and ~164 kcps at ~3.0 kBq/cm3 for the extended phantom. For an achievable value of τps = 1.0 μs, NECR was ~349 kcps (~486 kcps) at ~7.6 kBq/cm3 (~16.8 kBq/cm3) for the standard NEMA NU2-2001 phantom, and ~323 kcps (~460 kcps) at ~2.9 kBq/cm3 (~6.5 kBq/cm3) for the extended one. In conclusion, present and previous works reveals that RPC TOF-PET is expected to outperform current PET scanners.
Paulo Crespo, João Reis, Miguel Couceiro, Alberto Blanco, Nuno C. Ferreira, Rui Ferreira Marques, Paulo Martins, and Paulo Fonte
Institute of Electrical and Electronics Engineers (IEEE)
A single-bed, whole-body positron emission tomograph based on resistive plate chambers has been proposed (RPC-PET). An RPC-PET system with an axial field-of-view (AFOV) of 2.4 m has been shown in simulation to have higher system sensitivity using the NEMA NU2-1994 protocol than commercial PET scanners. However, that protocol does not correlate directly with lesion detectability. The latter is better correlated with the planar (slice) sensitivity, obtained with a NEMA NU2-2001 line-source phantom. After validation with published data for the GE Advance, Siemens TruePoint and TrueV, we study by simulation their axial sensitivity profiles, comparing results with RPC-PET. Planar sensitivities indicate that RPC-PET is expected to outperform 16-cm (22-cm) AFOV scanners by a factor 5.8 (3.0) for 70-cm-long scans. For 1.5-m scans (head to mid-legs), the sensitivity gain increases to 11.7 (6.7). Yet, PET systems with large AFOV provide larger coverage but also larger attenuation in the object. We studied these competing effects with both spherical- and line-sources immersed in a 27-cm-diameter water cylinder. For 1.5-m-long scans, the planar sensitivity drops one order of magnitude in all scanners, with RPC-PET outperforming 16-cm (22-cm) AFOV scanners by a factor 9.2 (5.3) without considering the TOF benefit. A gain in the effective sensitivity is expected with TOF iterative reconstruction. Finally, object scatter in an anthropomorphic phantom is similar for RPC-PET and modern, scintillator-based scanners, although RPC-PET benefits further if its TOF information is utilized to exclude scatter events occurring outside the anthropomorphic phantom.
M. Couceiro, P. Crespo, L. Mendes, N. Ferreira, R. Ferreira Marques, and P. Fonte
Elsevier BV
Abstract In this work, the spatial resolution of a human wide axial field of view RPC-PET system was studied. It was verified that physical processes involving the detection of 511 keV photons do not degrade the spatial resolution, compared with that attainable with systems based on inorganic scintillation crystals coupled to photomultiplier tubes. Considering the photon acollinearity effect, a detector depth-of-interaction resolution of 3.44 mm, and detector readout binnings of 1.0 and 2.0 mm, the spatial resolutions computed as the mean of full width at half maximum of point spread functions in three orthogonal directions (X, Y, and Z) were 1.4±0.0 and 2.1±0.1 mm, respectively. The corresponding mean spatial resolutions computed as the full width at tenth maximum were 3.9±0.4 and 4.8±0.3 mm, respectively.
Paulo Martins, Alberto Blanco, Francisco Caramelo, Miguel Couceiro, Nuno C. Ferreira, Rui Ferreira Marques, Paulo Fonte, Luis Mendes, and Paulo Crespo
IEEE
A single-bed, whole-body positron emission tomograph based on resistive plate chamber detectors has been proposed (RPC-PET). It has been shown by simulation that RPC-PET with an axial field-of-view (AFOV) of 2.4m is feasible and yields an absolute sensitivity enhancement of c. one order of magnitude superior to that of typical cylindrical, crystal-based PET scanners, with 16cm AFOV. In addition to its time-of-flight (TOF) advantage, RPC-PET offers potential very-high spatial resolution at the detector level. This must be properly handled by a fully-3D reconstruction algorithm capable of processing the very inclined lines-of-response (LOR) from large AFOV systems such as RPC-PET. For RPC-PET, these are acquired within a single-bed examination. Consequently, one or several segmented images representing the full human body during reconstruction are necessary. In this paper, we show that a direct-TOF implementation of the ordered subset expectation maximization (OSEM) algorithm allows for all events acquired with RPC-PET to be directly processed without rebinning and directly inserted inside the object image by means of a TOF-kernel. Such kernel (1) avoids typical (and slow) voxel-wise image navigation, (2) respects the time uncertainty dictated by the imaging RPC detectors, and (3) permits handling list-mode data iteratively with CT-based attenuation correction and OSEM computed both on-the-fly. We present reconstructed result from blind simulations of a single-bed, whole body PET system, corresponding to (1) six simulated spherical sources immersed in a homogeneous activity background, (2) an anthropomorphic phantom including a realistic grey-to-white matter uptake ratio of four, (3) and an additional anthropomorphic phantom with oncological lesions. We show that overall specificity for lesion detection may be increased (false positives decreased) by multiple similar, but independent reconstructions that eliminate false candidates arising from image statistical noise.
Paulo Crespo, Joao Reis, Miguel Couceiro, Alberto Blanco, Nuno C. Ferreira, Rui Ferreira Marques, and Paulo Fonte
IEEE
A single-bed, whole-body positron emission tomograph based on resistive plate chamber detectors has been proposed (RPC-PET). RPC-PET with an axial field-of-view (AFOV) of 2.4 m is feasible and yields an absolute NEMA NU 2-1994 sensitivity enhancement of 20 (4.5) in respect to 16-cm AFOV PET systems with (without) time-of-flight (TOF). These results, however, do not correlate directly with lesion detectability. It is the planar (slice) sensitivity that dictates the exposure time necessary to obtain enough statistics to detect a lesion. This planar sensitivity is currently obtained with a NEMA NU 2-2001 line-source phantom. After validating our simulations with measurements published for existing scanners (GE Advance, Siemens TruePoint and TrueV), we study here by simulation the axial sensitivity profiles of state-of-the-art BGO- and LSO-based PET scanners, and compare the results with RPC-PET. Planar sensitivity results indicate that RPC-PET is expected to outperform 16-cm AFOV scanners by a factor 6 for a 70-cm long scan. If the axial extent is elongated to 1.5 m (approximately head to mid-legs), the sensitivity gain increases to 12. Yet, PET systems with larger AFOV do provide a larger solid angle coverage, at the expense of larger attenuation in the object. In order to quantitate these competing effects, we have studied both point- and line-sources immersed in a water cylinder. For 1.5-m-long scans, the planar sensitivity drops over one order of magnitude in all scanners simulated. In this scenario, RPC-PET outperforms 16-cm AFOV scanners by a factor of 55 (9) with (without) considering the TOF benefit. Regarding the TOF benefit, RPC-PET has a sensitivity gain of 18 when compared to modern, 16-cm AFOV scanners with 600 ps FWHM TOF resolution. We finally show that object scatter in an anthropomorphic phantom is similar for a whole-body, single-bed RPC-PET in respect to a modern, scintillator-based scanner.
A. Blanco, M. Couceiro, P. Crespo, N.C. Ferreira, R. Ferreira Marques, P. Fonte, L. Lopes, and J.A. Neves
Elsevier BV
Abstract The resistive plate chamber (RPC) concept for time-of-flight positron emission tomography (TOF-PET) is based on the converter-plate gamma detection principle and takes advantage of the naturally layered structure of RPCs, of its simple and economic construction, excellent time resolution and very good intrinsic position accuracy. These characteristics may be of interest for the detailed imaging of small animals and for high-sensitivity whole-body human TOF-PET. In this communication we will present detailed simulations concerning the efficiency and imaging accuracy of the RPC-PET concept as a function of the main structural parameters, supplemented by measurements in a large-area chamber targeted at human RPC-PET.
M. Couceiro, A. Blanco, Nuno C. Ferreira, R. Ferreira Marques, P. Fonte, and L. Lopes
Elsevier BV
The status of the RPC-PET technology for small animals is briefly reviewed and its sensitivity performance for human PET studied through Monte-Carlo simulations. The cost-effectiveness of these detectors and their very good timing characteristics open the possibility to build affordable TOF-PET systems with very large fields of view. Simulations suggest that the sensitivity of such systems for human wholebody screening, under reasonable assumptions, may exceed the present crystal based PET technology by a factor up to 20.
M. Couceiro, N.C. Ferreira, and P. Fonte
Elsevier BV
Abstract The sensitivity characteristics of Positron Emission Tomography (PET) systems with wide Axial Field of View (AFOV) was studied by MonteCarlo simulations complemented by an approximate analytical model, aiming at full-body human PET systems with AFOV in the order of 200 cm. Simulations were based on the GEANT4 package and followed closely the NEMA NU-2 1994 norm. The sensitivity, dominated by the solid angle, grows strongly with the AFOV and with the axial acceptance angle, while the scatter fraction is almost independent from the geometry.