Fisher information-based evaluation of image quality for time-of-flight PET

Time of Flight in Perspective: Instrumental and Computational Aspects of Time Resolution in Positron Emission Tomography

The first time-of-flight positron emission tomography (TOF-PET) scanners were developed as early as in the 1980s. However, the poor light output and low detection efficiency of TOF-capable detectors available at the time limited any gain in image quality achieved with these TOF-PET scanners over the traditional non-TOF PET scanners. The discovery of LSO and other Lu-based scintillators revived interest in TOF-PET and led to the development of a second generation of scanners with high sensitivity and spatial resolution in the mid-2000s. The introduction of the silicon photomultiplier (SiPM) has recently yielded a third generation of TOF-PET systems with unprecedented imaging performance. Parallel to these instrumentation developments, much progress has been made in the development of image reconstruction algorithms that better utilize the additional information provided by TOF. Overall, the benefits range from a reduction in image variance (SNR increase), through allowing joint estimation of activity and attenuation, to better reconstructing data from limited angle systems. In this work, we review these developments, focusing on three broad areas: 1) timing theory and factors affecting the time resolution of a TOF-PET system; 2) utilization of TOF information for improved image reconstruction; and 3) quantification of the benefits of TOF compared to non-TOF PET. Finally, we offer a brief outlook on the TOF-PET developments anticipated in the short and longer term. Throughout this work, we aim to maintain a clinically driven perspective, treating TOF as one of multiple (and sometimes competitive) factors that can aid in the optimization of PET imaging performance.

Effects of TOF Resolution Models on Edge Artifacts in PET Reconstruction from Limited-Angle Data

Limited-angle data, such as data obtained from a dual-panel Breast-PET scanner, result in substantial image blur in directions coinciding with the missing cone of the image spectrum. On systems with time-of-flight (TOF) capabilities, this blur is reduced as given by the TOF uncertainty, with the image spectrum being correspondingly expanded into the missing spectral cone. Modeling of the TOF uncertainty in the reconstruction is expected to deconvolve this residual TOF blurring. We have however observed that, as a tradeoff, this TOF de-blurring process also introduces ringing artifacts at the edges, analogous to the edge effects observed with line-of-response (LOR) resolution modeling, which attempts to deconvolve the blur due to detector resolution effects. However, in the former case, the ringing artifacts are much wider due to the spatial extent of the TOF uncertainty as compared to the width of typical LOR resolution blur. We illustrate and investigate the effects of using matched, as well as under-modeled and over-modeled, TOF kernels on edge artifacts in reconstruction from limited-angle data, and compare them with TOF reconstructions of complete data. Although for the conventional data with full angular coverage the reconstruction is fairly insensitive to the exact size of the TOF kernel and TOF modeling does not produce ringing artifacts, it is not the case for the limited-angle data. We show that it is important to use some form of regularization of the TOF uncertainty deconvolution process within reconstruction of the limited-angle data, such as decreasing the TOF kernel size.

Estimation of Crystal Timing Properties and Efficiencies for the Improvement of (Joint) Maximum-Likelihood Reconstructions in TOF-PET

With increasing improvements in the time of flight (TOF) resolution of positron emission tomography (PET) scanners, an accurate model of the TOF measurements is becoming increasingly important. This work considers two parameters of the TOF kernel; the relative positioning of the timing data-bins and the timing resolution along each line of response (LOR). Similar to an existing data-driven method, we assume that any shifts of data-bins along lines of response can be modelled as differences between crystal timing offsets. Inspired by this, timing resolutions of all LORs are modelled as the hypotenuse of timing resolutions of the crystal-pairs in coincidence. Furthermore, in order to mitigate the influence of potential inaccuracies of detector-pair sensitivities on crystal timing resolutions, relative LOR sensitivities are modelled as the product of efficiency factors for the two crystals in coincidence. We validate estimating maps of crystal timing offsets, timing resolutions and efficiencies from the emission data using noisy simulations of a brain phantom. Results are shown for phantom and patient data scanned on clinically available TOF-PET scanners. We find that the estimation of crystal timing resolutions is more sensitive to the data statistics than the estimation of crystal timing offsets. As a result, estimation of crystal timing properties could either be limited to high count emission data, or be obtained utilizing additional regularizations on the estimates. Using a more accurate model of the TOF acquisition, improvements are observed in standard activity reconstructions as well as joint reconstructions of activity and attenuation.

Quantifying the loss of information from binning list-mode data

List-mode data are increasingly being used in single photon emission computed tomography (SPECT) and positron emission tomography (PET) imaging, among other imaging modalities. However, there are still many imaging designs that effectively bin list-mode data before image reconstruction or other estimation tasks are performed. Intuitively, the binning operation should result in a loss of information. In this work, we show that this is true for Fisher information and provide a computational method for quantifying the information loss. In the end, we find that the information loss depends on three factors. The first factor is related to the smoothness of the mean data function for the list-mode data. The second factor is the actual object being imaged. Finally, the third factor is the binning scheme in relation to the other two factors.

Time efficient scatter correction for time-of-flight PET: the immediate scatter approximation

Utilization of time-of-flight (TOF) information allows us to improve image quality and convergence rate in iterative PET image reconstruction. In order to obtain quantitatively correct images accurate scatter correction (SC) is required that accounts for the non-uniform distribution of scatter events over the TOF bins. However, existing simplified TOF-SC algorithms frequently exhibit limited accuracy while the currently accepted reference method-the TOF extension of the single scatter simulation approach (TOF-SSS)-is computationally demanding and can substantially slow down the reconstruction. In this paper we propose and evaluate a new accelerated TOF-SC algorithm in order to improve this situation. The key idea of the algorithm is the use of an immediate scatter approximation (ISA) for scatter time distribution calculation which speeds up estimation of the required TOF scatter by a factor of up to five in comparison to TOF-SSS. The proposed approach was evaluated in dedicated phantom measurements providing challenging high activity contrast conditions as well as in representative clinical patient data sets. Our results show that ISA is a viable alternative to TOF-SSS. The reconstructed images are in excellent quantitative agreement with those obtained with TOF-SSS while overall reconstruction time can be reduced by a factor of two in whole-body studies, even when using a listmode reconstruction not optimized for speed.

The origin and reduction of spurious extrahepatic counts observed in 90Y non-TOF PET imaging post radioembolization

Our literature survey revealed a physical effect unknown to the nuclear medicine community, i.e. internal bremsstrahlung emission, and also the existence of long energy resolution tails in crystal scintillation. None of these effects has ever been modelled in PET Monte Carlo (MC) simulations. This study investigates whether these two effects could be at the origin of two unexplained observations in ⁹⁰Y imaging by PET: the increasing tails in the radial profile of true coincidences, and the presence of spurious extrahepatic counts post radioembolization in non-TOF PET and their absence in TOF PET. These spurious extrahepatic counts hamper the microsphere delivery check in liver radioembolization. An acquisition of a ³²P vial was performed on a GSO PET system. This is the ideal setup to study the impact of bremsstrahlung x-rays on the true coincidence rate when no positron emission and no crystal radioactivity are present. A MC simulation of the acquisition was performed using Gate-Geant4. MC simulations of non-TOF PET and TOF-PET imaging of a synthetic ⁹⁰Y human liver radioembolization phantom were also performed. Internal bremsstrahlung and long energy resolution tails inclusion in MC simulations quantitatively predict the increasing tails in the radial profile. In addition, internal bremsstrahlung explains the discrepancy previously observed in bremsstrahlung SPECT between the measure of the ⁹⁰Y bremsstrahlung spectrum and its simulation with Gate-Geant4. However the spurious extrahepatic counts in non-TOF PET mainly result from the failure of conventional random correction methods in such low count rate studies and poor robustness versus emission-transmission inconsistency. A novel proposed random correction method succeeds in cleaning the spurious extrahepatic counts in non-TOF PET. Two physical effects not considered up to now in nuclear medicine were identified to be at the origin of the unusual ⁹⁰Y true coincidences radial profile. TOF reconstruction removing of the spurious extrahepatic counts was theoretically explained by a better robustness against emission-transmission inconsistency. A novel random correction method was proposed to overcome the issue in non-TOF PET. Further studies are needed to assess the novel random correction method robustness.

Lesion detection and quantification performance of the Tachyon-I time-of-flight PET scanner: phantom and human studies

The first generation Tachyon PET (Tachyon-I) is a demonstration single-ring PET scanner that reaches a coincidence timing resolution of 314 ps using LSO scintillator crystals coupled to conventional photomultiplier tubes. The objective of this study was to quantify the improvement in both lesion detection and quantification performance resulting from the improved time-of-flight (TOF) capability of the Tachyon-I scanner. We developed a quantitative TOF image reconstruction method for the Tachyon-I and evaluated its TOF gain for lesion detection and quantification. Scans of either a standard NEMA torso phantom or healthy volunteers were used as the normal background data. Separately scanned point source and sphere data were superimposed onto the phantom or human data after accounting for the object attenuation. We used the bootstrap method to generate multiple independent noisy datasets with and without a lesion present. The signal-to-noise ratio (SNR) of a channelized hotelling observer (CHO) was calculated for each lesion size and location combination to evaluate the lesion detection performance. The bias versus standard deviation trade-off of each lesion uptake was also calculated to evaluate the quantification performance. The resulting CHO-SNR measurements showed improved performance in lesion detection with better timing resolution. The detection performance was also dependent on the lesion size and location, in addition to the background object size and shape. The results of bias versus noise trade-off showed that the noise (standard deviation) reduction ratio was about 1.1-1.3 over the TOF 500 ps and 1.5-1.9 over the non-TOF modes, similar to the SNR gains for lesion detection. In conclusion, this Tachyon-I PET study demonstrated the benefit of improved time-of-flight capability on lesion detection and ROI quantification for both phantom and human subjects.

Monte Carlo simulations of time-of-flight PET with double-ended readout: calibration, coincidence resolving times and statistical lower bounds

This paper demonstrates through Monte Carlo simulations that a practical positron emission tomograph with (1) deep scintillators for efficient detection, (2) double-ended readout for depth-of-interaction information, (3) fixed-level analog triggering, and (4) accurate calibration and timing data corrections can achieve a coincidence resolving time (CRT) that is not far above the statistical lower bound. One Monte Carlo algorithm simulates a calibration procedure that uses data from a positron point source. Annihilation events with an interaction near the entrance surface of one scintillator are selected, and data from the two photodetectors on the other scintillator provide depth-dependent timing corrections. Another Monte Carlo algorithm simulates normal operation using these corrections and determines the CRT. A third Monte Carlo algorithm determines the CRT statistical lower bound by generating a series of random interaction depths, and for each interaction a set of random photoelectron times for each of the two photodetectors. The most likely interaction times are determined by shifting the depth-dependent probability density function to maximize the joint likelihood for all the photoelectron times in each set. Example calculations are tabulated for different numbers of photoelectrons and photodetector time jitters for three 3  ×  3  ×  30 mm³ scintillators: Lu₂SiO₅:Ce,Ca (LSO), LaBr₃:Ce, and a hypothetical ultra-fast scintillator. To isolate the factors that depend on the scintillator length and the ability to estimate the DOI, CRT values are tabulated for perfect scintillator-photodetectors. For LSO with 4000 photoelectrons and single photoelectron time jitter of the photodetector J  =  0.2 ns (FWHM), the CRT value using the statistically weighted average of corrected trigger times is 0.098 ns FWHM and the statistical lower bound is 0.091 ns FWHM. For LaBr₃:Ce with 8000 photoelectrons and J  =  0.2 ns FWHM, the CRT values are 0.070 and 0.063 ns FWHM, respectively. For the ultra-fast scintillator with 1 ns decay time, 4000 photoelectrons, and J  =  0.2 ns FWHM, the CRT values are 0.021 and 0.017 ns FWHM, respectively. The examples also show that calibration and correction for depth-dependent variations in pulse height and in annihilation and optical photon transit times are necessary to achieve these CRT values.

Monte Carlo calculations of PET coincidence timing: single and double-ended readout

We present Monte Carlo computational methods for estimating the coincidence resolving time (CRT) of scintillator detector pairs in positron emission tomography (PET) and present results for Lu2SiO5 : Ce (LSO), LaBr3 : Ce, and a hypothetical ultra-fast scintillator with a 1 ns decay time. The calculations were applied to both single-ended and double-ended photodetector readout with constant-fraction triggering. They explicitly include (1) the intrinsic scintillator properties (luminosity, rise time, decay time, and index of refraction), (2) the exponentially distributed depths of interaction, (3) the optical photon transport efficiency, delay, and time dispersion, (4) the photodetector properties (fill factor, quantum efficiency, transit time jitter, and single electron response), and (5) the determination of the constant fraction trigger level that minimizes the CRT. The calculations for single-ended readout include the delayed photons from the opposite reflective surface. The calculations for double-ended readout include (1) the simple average of the two photodetector trigger times, (2) more accurate estimators of the annihilation photon entrance time using the pulse height ratio to estimate the depth of interaction and correct for annihilation photon, optical photon, and trigger delays, and (3) the statistical lower bound for interactions at the center of the crystal. For time-of-flight (TOF) PET we combine stopping power and TOF information in a figure of merit equal to the sensitivity gain relative to whole-body non-TOF PET using LSO. For LSO crystals 3 mm  ×  3 mm  ×  30 mm, a decay time of 37 ns, a total photoelectron count of 4000, and a photodetector with 0.2 ns full-width at half-maximum (fwhm) timing jitter, single-ended readout has a CRT of 0.16 ns fwhm and double-ended readout has a CRT of 0.111 ns fwhm. For LaBr3 : Ce crystals 3 mm  ×  3 mm  ×  30 mm, a rise time of 0.2 ns, a decay time of 18 ns, and a total of 7600 photoelectrons the CRT numbers are 0.14 ns and 0.072 ns fwhm, respectively. For a hypothetical ultra-fast scintillator 3 mm  ×  3 mm  ×  30 mm, a decay time of 1 ns, and a total of 4000 photoelectrons, the CRT numbers are 0.070 and 0.020 ns fwhm, respectively. Over a range of examples, values for double-ended readout are about 10% larger than the statistical lower bound.

Efficient determination of the uncertainty for the optimization of SPECT system design: a subsampled fisher information matrix

System designs in single photon emission tomography (SPECT) can be evaluated based on the fundamental trade-off between bias and variance that can be achieved in the reconstruction of emission tomograms. This trade off can be derived analytically using the Cramer-Rao type bounds, which imply the calculation and the inversion of the Fisher information matrix (FIM). The inverse of the FIM expresses the uncertainty associated to the tomogram, enabling the comparison of system designs. However, computing, storing and inverting the FIM is not practical with 3-D imaging systems. In order to tackle the problem of the computational load in calculating the inverse of the FIM, a method based on the calculation of the local impulse response and the variance, in a single point, from a single row of the FIM, has been previously proposed for system design. However this approximation (circulant approximation) does not capture the global interdependence between the variables in shift-variant systems such as SPECT, and cannot account e.g., for data truncation or missing data. Our new formulation relies on subsampling the FIM. The FIM is calculated over a subset of voxels arranged in a grid that covers the whole volume. Every element of the FIM at the grid points is calculated exactly, accounting for the acquisition geometry and for the object. This new formulation reduces the computational complexity in estimating the uncertainty, but nevertheless accounts for the global interdependence between the variables, enabling the exploration of design spaces hindered by the circulant approximation. The graphics processing unit accelerated implementation of the algorithm reduces further the computation times, making the algorithm a good candidate for real-time optimization of adaptive imaging systems. This paper describes the subsampled FIM formulation and implementation details. The advantages and limitations of the new approximation are explored, in comparison with the circulant approximation, in the context of design optimization of a parallel-hole collimator SPECT system and of an adaptive imaging system (similar to the commercially available D-SPECT).

MAP reconstruction for Fourier rebinned TOF-PET data

Time-of-flight (TOF) information improves the signal-to-noise ratio in positron emission tomography (PET). The computation cost in processing TOF-PET sinograms is substantially higher than for nonTOF data because the data in each line of response is divided among multiple TOF bins. This additional cost has motivated research into methods for rebinning TOF data into lower dimensional representations that exploit redundancies inherent in TOF data. We have previously developed approximate Fourier methods that rebin TOF data into either three-dimensional (3D) nonTOF or 2D nonTOF formats. We refer to these methods respectively as FORET-3D and FORET-2D. Here we describe maximum a posteriori (MAP) estimators for use with FORET rebinned data. We first derive approximate expressions for the variance of the rebinned data. We then use these results to rescale the data so that the variance and mean are approximately equal allowing us to use the Poisson likelihood model for MAP reconstruction. MAP reconstruction from these rebinned data uses a system matrix in which the detector response model accounts for the effects of rebinning. Using these methods we compare the performance of FORET-2D and 3D with TOF and nonTOF reconstructions using phantom and clinical data. Our phantom results show a small loss in contrast recovery at matched noise levels using FORET compared to reconstruction from the original TOF data. Clinical examples show FORET images that are qualitatively similar to those obtained from the original TOF-PET data but with a small increase in variance at matched resolution. Reconstruction time is reduced by a factor of 5 and 30 using FORET3D+MAP and FORET2D+MAP respectively compared to 3D TOF MAP, which makes these methods attractive for clinical applications.

Simultaneous reconstruction of activity and attenuation in time-of-flight PET

In positron emission tomography (PET) and single photon emission tomography (SPECT), attenuation correction is necessary for quantitative reconstruction of the tracer distribution. Previously, several attempts have been made to estimate the attenuation coefficients from emission data only. These attempts had limited success, because the problem does not have a unique solution, and severe and persistent "cross-talk" between the estimated activity and attenuation distributions was observed. In this paper, we show that the availability of time-of-flight (TOF) information eliminates the cross-talk problem by destroying symmetries in the associated Fisher information matrix. We propose a maximum-a-posteriori reconstruction algorithm for jointly estimating the attenuation and activity distributions from TOF PET data. The performance of the algorithm is studied with 2-D simulations, and further illustrated with phantom experiments and with a patient scan. The estimated attenuation image is robust to noise, and does not suffer from the cross-talk that was observed in non-TOF PET. However, some constraining is still mandatory, because the TOF data determine the attenuation sinogram only up to a constant offset.

Optimal rebinning of time-of-flight PET data

Time-of-flight (TOF) positron emission tomography (PET) scanners offer the potential for significantly improved signal-to-noise ratio (SNR) and lesion detectability in clinical PET. However, fully 3D TOF PET image reconstruction is a challenging task due to the huge data size. One solution to this problem is to rebin TOF data into a lower dimensional format. We have recently developed Fourier rebinning methods for mapping TOF data into non-TOF formats that retain substantial SNR advantages relative to sinograms acquired without TOF information. However, mappings for rebinning into non-TOF formats are not unique and optimization of rebinning methods has not been widely investigated. In this paper we address the question of optimal rebinning in order to make full use of TOF information. We focus on FORET-3D, which approximately rebins 3D TOF data into 3D non-TOF sinogram formats without requiring a Fourier transform in the axial direction. We optimize the weighting for FORET-3D to minimize the variance, resulting in H(2)-weighted FORET-3D, which turns out to be the best linear unbiased estimator (BLUE) under reasonable approximations and furthermore the uniformly minimum variance unbiased (UMVU) estimator under Gaussian noise assumptions. This implies that any information loss due to optimal rebinning is as a result only of the approximations used in deriving the rebinning equation and developing the optimal weighting. We demonstrate using simulated and real phantom TOF data that the optimal rebinning method achieves variance reduction and contrast recovery improvement compared to nonoptimized rebinning weightings. In our preliminary study using a simplified simulation setup, the performance of the optimal rebinning method was comparable to that of fully 3D TOF MAP.

Comparison between parallel hole and rotating slat collimation: analytical noise propagation models

We have previously proposed a method to compare tomographic systems. It is assumed that each system acquires a tomographic scan of a certain tracer distribution in the same acquisition time. From this scan, each system is forced to reconstruct an image with a predefined spatial resolution. The system that can perform this task with the "most favorable" noise propagation is considered as the best system. The variance on pixel values or region-of-interest (ROI) values is used to assess the noise in the reconstructed image. In this paper, we extend this idea to compare the performance of parallel hole (PH) and rotating slat (RS) collimations. Two different analytical approaches were used to analyze the variance of the reconstructed pixel/ROI values. The first method is based on the filtered-backprojection (FBP) theory, and was applied to the central point of a uniform symmetrical phantom. It yields analytical expressions for the optimal collimator aperture and the corresponding variance of the reconstructed pixel values, but it can only be applied to highly symmetrical configurations. The second method is based on approximations for the Fisher information matrix. It provides numerical results, and it is more general and can be applied to nonsymmetrical objects and shift-variant tomographic systems. The collimations were compared for both planar imaging and volume imaging. The main results are as follows. 1) For cases where both methods are valid, they are in excellent agreement. 2a) The optimal collimator aperture varies linearly with the target resolution. 2b) For a fixed target resolution, the optimal collimator aperture depends on the collimator type and the imaging mode (planar or volume). 2c) The optimal aperture of PH is a factor of √2 larger than that of RS. 3a) The relative performance of the two collimators is determined by both the object size and the object-to-detector distance. 3b) Pixel variance and variances of ROIs with varying sizes yield very similar relative performance for RS versus PH.

Detection performance analysis for time-of-flight PET

In this paper, we investigate the performance of time-of-flight (TOF) positron emission tomography (PET) in improving lesion detectability. We present a theoretical approach to compare lesion detectability of TOF versus non-TOF systems and perform computer simulations to validate the theoretical prediction. A single-ring TOF PET tomograph is simulated using SimSET software, and images are reconstructed in 2D from list-mode data using a maximum a posteriori method. We use a channelized Hotelling observer to assess the detection performance. Both the receiver operating characteristic (ROC) and localization ROC curves are compared for the TOF and non-TOF PET systems. We first studied the SNR gains for TOF PET with different scatter and random fractions, system timing resolutions and object sizes. We found that the TOF information improves the lesion detectability and the improvement is greater with larger fractions of randoms, better timing resolution and bigger objects. The scatters by themselves have little impact on the SNR gain after correction. Since the true system timing resolution may not be known precisely in practice, we investigated the effect of mismatched timing kernels and showed that using a mismatched kernel during reconstruction always degrades the detection performance, no matter whether it is narrower or wider than the real value. Using the proposed theoretical framework, we also studied the effect of lumpy backgrounds on the detection performance. Our results indicated that with lumpy backgrounds, the TOF PET still outperforms the non-TOF PET, but the improvement is smaller compared with the uniform background case. More specifically, with the same correlation length, the SNR gain reduces with bigger number of lumpy patches and greater lumpy amplitudes. With the same variance, the SNR gain reaches the minimum when the width of the Gaussian lumps is close to the size of the tumor.