Design of a generic method for single dual-tracer PET imaging acquisition in clinical routine.

Using different tracers in positron emission tomography (PET) imaging can bring complementary information on tumor heterogeneities. Ideally, PET images of different tracers should be acquired simultaneously to avoid the bias induced by movement and physiological changes between sequential acquisitions. Previous studies have demonstrated the feasibility of recovering separated PET signals or parameters of two or more tracers injected (quasi-)simultaneously in a single acquisition. In this study, a generic framework in the context of dual-tracer PET acquisition is proposed where no strong kinetic assumptions nor specific tuning of parameters are required. The performances of the framework were assessed through simulations involving the combination of [18F]FCH and [18F]FDG injections, two protocols (90 and 60 min acquisition durations) and various activity ratios between the two injections. Preclinical experiments with the same radiotracers were also conducted. Results demonstrate the ability of the method both to extract separated arterial input functions (AIF) from noisy image-derived input function and to separate the dynamic signals and further estimate kinetic parameters. The compromise between bias and variance associated with the estimation of net influx rateKishows that it is preferable to use the second injected radiotracer with twice the activity of the first for both 90 min [18F]FCH+[18F]FDG and 60 min [18F]FDG+[18F]FCH protocols. In these optimal settings, the weighted mean-squared-error of the estimated AIF was always less than 7%. TheKibias was similar to the one of single-tracer acquisitions; below 5%. Compared to single-tracer results, the variance ofKiwas twice more for 90 min dual-tracer scenario and four times more for the 60 min scenario. The generic design of the method makes it easy to use for other pairs of radiotracers and even for more than two tracers. The absence of strong kinetic assumptions and tuning parameters makes it suitable for a possible use in clinical routine.

Use of dynamic reconstruction for parametric Patlak imaging in dynamic whole body PET.

Dynamic whole body (DWB) PET acquisition protocols enable the use of whole body parametric imaging for clinical applications. In FDG imaging, accurate parametric images of PatlakKican be complementary to regular standardised uptake value images and improve on current applications or enable new ones. In this study we consider DWB protocols implemented on clinical scanners with a limited axial field of view with the use of multiple whole body sweeps. These protocols result in temporal gaps in the dynamic data which produce noisier and potentially more biased parametric images, compared to single bed (SB) dynamic protocols. Dynamic reconstruction using the Patlak model has been previously proposed to overcome these limits and shown improved DWB parametric images ofKi. In this work, we propose and make use of a spectral analysis based model for dynamic reconstruction and parametric imaging of PatlakKi. Both dynamic reconstruction methods were evaluated for DWB FDG protocols and compared against 3D reconstruction based parametric imaging from SB dynamic protocols. This work was conducted on simulated data and results were tested against real FDG dynamic data. We showed that dynamic reconstruction can achieve levels of parametric image noise and bias comparable to 3D reconstruction in SB dynamic studies, with the spectral model offering additional flexibility and further reduction of image noise. Comparisons were also made between step and shoot and continuous bed motion (CBM) protocols, which showed that CBM can achieve lower parametric image noise due to reduced acquisition temporal gaps. Finally, our results showed that dynamic reconstruction improved VOI parametric mean estimates but did not result to fully converged values before resulting in undesirable levels of noise. Additional regularisation methods need to be considered for DWB protocols to ensure both accurate quantification and acceptable noise levels for clinical applications.

Reconstruction, analysis and interpretation of posterior probability distributions of PET images, using the posterior bootstrap.

The uncertainty of reconstructed PET images remains difficult to assess and to interpret for the use in diagnostic and quantification tasks. Here we provide (1) an easy-to-use methodology for uncertainty assessment for almost any Bayesian model in PET reconstruction from single datasets and (2) a detailed analysis and interpretation of produced posterior image distributions. We apply a recent posterior bootstrap framework to the PET image reconstruction inverse problem and obtain simple parallelizable algorithms based on random weights and on existing maximuma posteriori(MAP) (posterior maximum) optimization-based algorithms. Posterior distributions are produced, analyzed and interpreted for several common Bayesian models. Their relationship with the distribution of the MAP image estimate over multiple dataset realizations is exposed. The coverage properties of posterior distributions are validated. More insight is obtained for the interpretation of posterior distributions in order to open the way for including uncertainty information into diagnostic and quantification tasks.

Introducing improved voxel navigation and fictitious interaction tracking in GATE for enhanced efficiency.

Geant4 Application for Emission Tomography (GATE) is a widely used, well-validated and very versatile application for Monte Carlo simulations in emission tomography. However, its computational performance is poor, especially for voxelized phantoms, partly due to the use of a very general particle tracking algorithm. In this work, two methods are proposed to reduce the time spent on particle tracking in the phantom: a newly introduced 'regular navigation algorithm' of Geant4 and fictitious interaction tracking (also known as Woodcock tracking) for photons. The speed-up introduced by the two methods was investigated by simulating a PET acquisition with the Allegro/GEMINI GXL PET/CT scanner. The simulation was based on a clinical head-and-neck [(18)F]FDG PET/CT scan. The total time spent for the simulation (including initialization, particle tracking and signal processing) was obtained for seven settings corresponding to different tracking options. All seven methods led to very close results with regard to the total number of detected coincidences (less than 0.5% differences), and trues, scatter and random fractions. Acceleration factors of approximately 2.7 (14 x 14 x 9 voxels) to 27.6 (378 x 378 x 243 voxels) were obtained in comparison with the fastest available tracking available in GATE 3.1.2.

Time-of-flight (TOF) implementation for PET reconstruction in practice.

The time-of-flight (TOF) feature of PET scanners has been used for a long time in PET reconstruction, but many implementational aspects are still incomplete or ambiguous in the literature. Here we formalize and present theoretical and practical implementation details for the reconstruction of clinical TOF histogram and list-mode data using ML-EM. Relevant aspects include the computation of the TOF component of the system matrix, the processing of TOF bins, the use of estimations of random and scattered coincidences, and differences between histogram and list-mode ML-EM TOF reconstruction. Several approaches and approximations have been implemented in the CASToR platform and compared for OSEM reconstructions of patient data from the GE Signa PET/MR scanner. Differences between implementations are not larger than the typical bias in clinical data reconstruction. The largest difference and contrast loss occur when the processing of histogram TOF bins is simplified, and list-mode reconstruction is most sensitive to the truncation of the Gaussian TOF probability distribution.

PET Reconstruction of the Posterior Image Probability, Including Multimodal Images.

In PET image reconstruction, it would be useful to obtain the entire posterior probability distribution of the image, because it allows for both estimating image intensity and assessing the uncertainty of the estimation, thus leading to more reliable interpretation. We propose a new entirely probabilistic model: the prior is a distribution over possible smooth regions (distance-driven Chinese restaurant process), and the posterior distribution is estimated using a Gibbs Markov chain Monte Carlo sampler. Data from other modalities (here one or several MR images) are introduced into the model as additional observed data, providing side information about likely smooth regions in the image. The reconstructed image is the posterior mean, and the uncertainty is presented as an image of the size of 95% posterior intervals. The reconstruction was compared with the maximum-likelihood expectation-maximization and OSEM algorithms, with and without post-smoothing, and with a penalized ML or MAP method that also uses additional images from other modalities. Qualitative and quantitative tests were performed on realistic simulated data with statistical replicates and on several clinical examinations presenting pathologies. The proposed method presents appealing properties in terms of obtained bias, variance, spatial regularization, and use of multimodal data, and produces, in addition, potentially valuable uncertainty information.

Factor Analysis of Dynamic PET Images: Beyond Gaussian Noise.

Factor analysis has proven to be a relevant tool for extracting tissue time-activity curves (TACs) in dynamic PET images, since it allows for an unsupervised analysis of the data. Reliable and interpretable results are possible only if it is considered with respect to suitable noise statistics. However, the noise in reconstructed dynamic PET images is very difficult to characterize, despite the Poissonian nature of the count rates. Rather than explicitly modeling the noise distribution, this paper proposes to study the relevance of several divergence measures to be used within a factor analysis framework. To this end, the ß -divergence, widely used in other applicative domains, is considered to design the data-fitting term involved in three different factor models. The performances of the resulting algorithms are evaluated for different values of ß , in a range covering Gaussian, Poissonian, and Gamma-distributed noises. The results obtained on two different types of synthetic images and one real image show the interest of applying non-standard values of ß to improve the factor analysis.

Unmixing dynamic PET images with variable specific binding kinetics.

To analyze dynamic positron emission tomography (PET) images, various generic multivariate data analysis techniques have been considered in the literature, such as principal component analysis (PCA), independent component analysis (ICA), factor analysis and nonnegative matrix factorization (NMF). Nevertheless, these conventional approaches neglect any possible nonlinear variations in the time activity curves describing the kinetic behavior of tissues with specific binding, which limits their ability to recover a reliable, understandable and interpretable description of the data. This paper proposes an alternative analysis paradigm that accounts for spatial fluctuations in the exchange rate of the tracer between a free compartment and a specifically bound ligand compartment. The method relies on the concept of linear unmixing, usually applied on the hyperspectral domain, which combines NMF with a sum-to-one constraint that ensures an exhaustive description of the mixtures. The spatial variability of the signature corresponding to the specific binding tissue is explicitly modeled through a perturbed component. The performance of the method is assessed on both synthetic and real data and is shown to compete favorably when compared to other conventional analysis methods.

CASToR: a generic data organization and processing code framework for multi-modal and multi-dimensional tomographic reconstruction.

In tomographic medical imaging (PET, SPECT, CT), differences in data acquisition and organization are a major hurdle for the development of tomographic reconstruction software. The implementation of a given reconstruction algorithm is usually limited to a specific set of conditions, depending on the modality, the purpose of the study, the input data, or on the characteristics of the reconstruction algorithm itself. It causes restricted or limited use of algorithms, differences in implementation, code duplication, impractical code development, and difficulties for comparing different methods. This work attempts to address these issues by proposing a unified and generic code framework for formatting, processing and reconstructing acquired multi-modal and multi-dimensional data. The proposed iterative framework processes in the same way elements from list-mode (i.e. events) and histogrammed (i.e. sinogram or other bins) data sets. Each element is processed separately, which opens the way for highly parallel execution. A unique iterative algorithm engine makes use of generic core components corresponding to the main parts of the reconstruction process. Features that are specific to different modalities and algorithms are embedded into specific components inheriting from the generic abstract components. Temporal dimensions are taken into account in the core architecture. The framework is implemented in an open-source C++ parallel platform, called CASToR (customizable and advanced software for tomographic reconstruction). Performance assessments show that the time loss due to genericity remains acceptable, being one order of magnitude slower compared to a manufacturer's software optimized for computational efficiency for a given system geometry. Specific optimizations were made possible by the underlying data set organization and processing and allowed for an average speed-up factor ranging from 1.54 to 3.07 when compared to more conventional implementations. Using parallel programming, an almost linear speed-up increase (factor of 0.85 times number of cores) was obtained in a realistic clinical PET setting. In conclusion, the proposed framework offers a substantial flexibility for the integration of new reconstruction algorithms while maintaining computation efficiency.

Imaging the Impact of the P-Glycoprotein (ABCB1) Function on the Brain Kinetics of Metoclopramide.

UNLABELLED: The effects of metoclopramide on the central nervous system (CNS) in patients suggest substantial brain distribution. Previous data suggest that metoclopramide brain kinetics may nonetheless be controlled by ATP-binding cassette (ABC) transporters expressed at the blood-brain barrier. We used (11)C-metoclopramide PET imaging to elucidate the kinetic impact of transporter function on metoclopramide exposure to the brain. METHODS: (11)C-metoclopramide transport by P-glycoprotein (P-gp; ABCB1) and the breast cancer resistance protein (BCRP; ABCG2) was tested using uptake assays in cells overexpressing P-gp and BCRP. (11)C-metoclopramide brain kinetics were compared using PET in rats (n = 4-5) in the absence and presence of a pharmacologic dose of metoclopramide (3 mg/kg), with or without P-gp inhibition using intravenous tariquidar (8 mg/kg). The (11)C-metoclopramide brain distribution (VT based on Logan plot analysis) and brain kinetics (2-tissue-compartment model) were characterized with either a measured or an imaged-derived input function. Plasma and brain radiometabolites were studied using radio-high-performance liquid chromatography analysis. RESULTS: (11)C-metoclopramide transport was selective for P-gp over BCRP. Pharmacologic dose did not affect baseline (11)C-metoclopramide brain kinetics (VT = 2.28 ± 0.32 and 2.04 ± 0.19 mL⋅cm(-3) using microdose and pharmacologic dose, respectively). Tariquidar significantly enhanced microdose (11)C-metoclopramide VT (7.80 ± 1.43 mL⋅cm(-3)) with a 4.4-fold increase in K1 (influx rate constant) and a 2.3-fold increase in binding potential (k3/k4) in the 2-tissue-compartment model. In the pharmacologic situation, P-gp inhibition significantly increased metoclopramide brain distribution (VT = 6.28 ± 0.48 mL⋅cm(-3)) with a 2.0-fold increase in K1 and a 2.2-fold decrease in k2 (efflux rate), with no significant impact on binding potential. In this situation, only parent (11)C-metoclopramide could be detected in the brains of P-gp-inhibited rats. CONCLUSION: (11)C-metoclopramide benefits from favorable pharmacokinetic properties that offer reliable quantification of P-gp function at the blood-brain barrier in a pharmacologic situation. Using metoclopramide as a model of CNS drug, we demonstrated that P-gp function not only reduces influx but also mediates the efflux from the brain back to the blood compartment, with additional impact on brain distribution. This PET-based strategy of P-gp function investigation may provide new insight on the contribution of P-gp to the variability of response to CNS drugs between patients.

Bias reduction for low-statistics PET: maximum likelihood reconstruction with a modified Poisson distribution.

Positron emission tomography data are typically reconstructed with maximum likelihood expectation maximization (MLEM). However, MLEM suffers from positive bias due to the non-negativity constraint. This is particularly problematic for tracer kinetic modeling. Two reconstruction methods with bias reduction properties that do not use strict Poisson optimization are presented and compared to each other, to filtered backprojection (FBP), and to MLEM. The first method is an extension of NEGML, where the Poisson distribution is replaced by a Gaussian distribution for low count data points. The transition point between the Gaussian and the Poisson regime is a parameter of the model. The second method is a simplification of ABML. ABML has a lower and upper bound for the reconstructed image whereas AML has the upper bound set to infinity. AML uses a negative lower bound to obtain bias reduction properties. Different choices of the lower bound are studied. The parameter of both algorithms determines the effectiveness of the bias reduction and should be chosen large enough to ensure bias-free images. This means that both algorithms become more similar to least squares algorithms, which turned out to be necessary to obtain bias-free reconstructions. This comes at the cost of increased variance. Nevertheless, NEGML and AML have lower variance than FBP. Furthermore, randoms handling has a large influence on the bias. Reconstruction with smoothed randoms results in lower bias compared to reconstruction with unsmoothed randoms or randoms precorrected data. However, NEGML and AML yield both bias-free images for large values of their parameter.

Variational segmentation of vector-valued images with gradient vector flow.

In this paper, we extend the gradient vector flow field for robust variational segmentation of vector-valued images. Rather than using scalar edge information, we define a vectorial edge map derived from a weighted local structure tensor of the image that enables the diffusion of the gradient vectors in accurate directions through the 4D gradient vector flow equation. To reduce the contribution of noise in the structure tensor, image channels are weighted according to a blind estimator of contrast. The method is applied to biological volume delineation in dynamic PET imaging, and validated on realistic Monte Carlo simulations of numerical phantoms as well as on real images.

Practical considerations for image-based PSF and blobs reconstruction in PET.

Iterative reconstructions in positron emission tomography (PET) need a model relating the recorded data to the object/patient being imaged, called the system matrix (SM). The more realistic this model, the better the spatial resolution in the reconstructed images. However, a serious concern when using a SM that accurately models the resolution properties of the PET system is the undesirable edge artefact, visible through oscillations near sharp discontinuities in the reconstructed images. This artefact is a natural consequence of solving an ill-conditioned inverse problem, where the recorded data are band-limited. In this paper, we focus on practical aspects when considering image-based point-spread function (PSF) reconstructions. To remove the edge artefact, we propose to use a particular case of the method of sieves (Grenander 1981 Abstract Inference New York: Wiley), which simply consists in performing a standard PSF reconstruction, followed by a post-smoothing using the PSF as the convolution kernel. Using analytical simulations, we investigate the impact of different reconstruction and PSF modelling parameters on the edge artefact and its suppression, in the case of noise-free data and an exactly known PSF. Using Monte-Carlo simulations, we assess the proposed method of sieves with respect to the choice of the geometric projector and the PSF model used in the reconstruction. When the PSF model is accurately known, we show that the proposed method of sieves succeeds in completely suppressing the edge artefact, though after a number of iterations higher than typically used in practice. When applying the method to realistic data (i.e. unknown true SM and noisy data), we show that the choice of the geometric projector and the PSF model does not impact the results in terms of noise and contrast recovery, as long as the PSF has a width close to the true PSF one. Equivalent results were obtained using either blobs or voxels in the same conditions (i.e. the blob's density function being the same as the voxel-based PSF). From a practical point-of-view, the method can be used to perform fast reconstructions based on very simple models (compared to sinogram-based PSF modelling), producing artefact-free images with a better compromise between noise and spatial resolution than images reconstructed without or with under-estimated PSF. Besides, the method inherently limits the spatial resolution in the reconstructed images to the intrinsic one of the PET system.

PET-based dose delivery verification in proton therapy: a GATE based simulation study of five PET system designs in clinical conditions.

PET is a promising technique for in vivo treatment verification in hadrontherapy. Three main PET geometries dedicated to in-beam treatment monitoring have been proposed in the literature: the dual-head PET geometry, the OpenPET geometry and the slanted-closed ring geometry. The aim of this work is to characterize the performance of two of these dedicated PET detectors in realistic clinical conditions. Several configurations of the dual-head PET and OpenPET systems were simulated using GATE v6.2. For the dual-head configuration, two aperture angles (15° and 45°) were studied. For the OpenPET system, two gaps between rings were investigated (110 and 160 mm). A full-ring PET system was also simulated as a reference. After preliminary evaluation of the sensitivity and spatial resolution using a Derenzo phantom, a real small-field head and neck treatment plan was simulated, with and without introducing patient displacements. No wash-out was taken into account. 3D maps of the annihilation photon locations were deduced from the PET data acquired right after the treatment session (5 min acquisition) using a dedicated OS-EM reconstruction algorithm. Detection sensitivity at the center of the field-of-view (FOV) varied from 5.2% (45° dual-head system) to 7.0% (full-ring PET). The dual-head systems had a more uniform efficiency within the FOV than the OpenPET systems. The spatial resolution strongly depended on the location within the FOV for the ϕ = 45° dual-head system and for the two OpenPET systems. All investigated architectures identified the magnitude of mispositioning introduced in the simulations within a 1.5 mm accuracy. The variability on the estimated mispositionings was less than 2 mm for all PET systems.

A new method for volume segmentation of PET images, based on possibility theory.

18F-fluorodeoxyglucose positron emission tomography (18FDG PET) has become an essential technique in oncology. Accurate segmentation and uptake quantification are crucial in order to enable objective follow-up, the optimization of radiotherapy planning, and therapeutic evaluation. We have designed and evaluated a new, nearly automatic and operator-independent segmentation approach. This incorporated possibility theory, in order to take into account the uncertainty and inaccuracy inherent in the image. The approach remained independent of PET facilities since it did not require any preliminary calibration. Good results were obtained from phantom images [percent error =18.38% (mean) ± 9.72% (standard deviation)]. Results on simulated and anatomopathological data sets were quantified using different similarity measures and showed the method was efficient (simulated images: Dice index =82.18% ± 13.53% for SUV =2.5 ). The approach could, therefore, be an efficient and robust tool for uptake volume segmentation, and lead to new indicators for measuring volume of interest activity.

Multidimensional B-spline parameterization of the detection probability of PET systems to improve the efficiency of Monte Carlo simulations.

Accurate modeling of system response and scatter distribution is crucial for image reconstruction in emission tomography. Monte Carlo simulations are very well suited to calculate these quantities. However, Monte Carlo simulations are also slow and many simulated counts are needed to provide a sufficiently exact estimate of the detection probabilities. In order to overcome these problems, we propose to split the simulation into two parts, the detection system and the object to be imaged (the patient). A so-called 'virtual boundary' that separates these two parts is introduced. Within the patient, particles are simulated conventionally. Whenever a photon reaches the virtual boundary, its detection probability is calculated analytically by evaluating a multi-dimensional B-spline that depends on the photon position, direction and energy. The unknown B-spline knot values that define this B-spline are fixed by a prior 'pre-' simulation that needs to be run once for each scanner type. After this pre-simulation, the B-spline model can be used in any subsequent simulation with different patients. We show that this approach yields accurate results when simulating the Biograph 16 HiREZ PET scanner with Geant4 Application for Emission Tomography (GATE). The execution time is reduced by a factor of about 22 x (scanner with voxelized phantom) to 30 x (empty scanner) with respect to conventional GATE simulations of same statistical uncertainty. The pre-simulation and calculation of the B-spline knots values could be performed within half a day on a medium-sized cluster.

Comparative assessment of methods for estimating tumor volume and standardized uptake value in (18)F-FDG PET.

UNLABELLED: In (18)F-FDG PET, tumors are often characterized by their metabolically active volume and standardized uptake value (SUV). However, many approaches have been proposed to estimate tumor volume and SUV from (18)F-FDG PET images, none of them being widely agreed upon. We assessed the accuracy and robustness of 5 methods for tumor volume estimates and of 10 methods for SUV estimates in a large variety of configurations. METHODS: PET acquisitions of an anthropomorphic phantom containing 17 spheres (volumes between 0.43 and 97 mL, sphere-to-surrounding-activity concentration ratios between 2 and 68) were used. Forty-one nonspheric tumors (volumes between 0.6 and 92 mL, SUV of 2, 4, and 8) were also simulated and inserted in a real patient (18)F-FDG PET scan. Four threshold-based methods (including one, T(bgd), accounting for background activity) and a model-based method (Fit) described in the literature were used for tumor volume measurements. The mean SUV in the resulting volumes were calculated, without and with partial-volume effect (PVE) correction, as well as the maximum SUV (SUV(max)). The parameters involved in the tumor segmentation and SUV estimation methods were optimized using 3 approaches, corresponding to getting the best of each method or testing each method in more realistic situations in which the parameters cannot be perfectly optimized. RESULTS: In the phantom and simulated data, the T(bgd) and Fit methods yielded the most accurate volume estimates, with mean errors of 2% +/- 11% and -8% +/- 21% in the most realistic situations. Considering the simulated data, all SUV not corrected for PVE had a mean bias between -31% and -46%, much larger than the bias observed with SUV(max) (-11% +/- 23%) or with the PVE-corrected SUV based on T(bgd) and Fit (-2% +/- 10% and 3% +/- 24%). CONCLUSION: The method used to estimate tumor volume and SUV greatly affects the reliability of the estimates. The T(bgd) and Fit methods yielded low errors in volume estimates in a broad range of situations. The PVE-corrected SUV based on T(bgd) and Fit were more accurate and reproducible than SUV(max).