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.

Convergence of iterative algorithms can be improved by updating groups of voxels sequentially rather than updating the whole image simultaneously. The optimal way is to choose groups of uncoupled voxels, i.e. voxels spread over the reconstruction volume. While this is most efficient for convergence reasons, updating groups of spread voxels is less efficient regarding memory access and computational burden. In this work, an image-block update scheme is presented that updates relatively large groups of voxels simultaneously while keeping a considerable gain in convergence. The sequential image-block update can also be combined with ordered subsets. This image-block or patchwork scheme is applied both to transmission and emission maximum likelihood algorithms.

Purpose: Digital breast tomosynthesis is a relatively new diagnostic x‐ray modality that allows high resolution breast imaging while suppressing interference from overlapping anatomical structures. However, proper visualization of microcalcifications remains a challenge. For the subset of systems considered by the authors, the main cause of deterioration is movement of the x‐ray source during exposures. They propose a modified grouped coordinate ascent algorithm that includes a specific acquisition model to compensate for this deterioration. Methods: A resolution model based on the movement of the x‐ray source during image acquisition is created and combined with a grouped coordinate ascent algorithm. Choosing planes parallel to the detector surface as the groups enables efficient implementation of the position dependent resolution model. In the current implementation, the resolution model is approximated by a Gaussian smoothing kernel. The effect of the resolution model on the iterative reconstruction is evaluated by measuring contrast to noise ratio (CNR) of spherical microcalcifications in a homogeneous background. After this, the new reconstruction method is compared to the optimized filtered backprojection method for the considered system, by performing two observer studies: the first study simulates clusters of spherical microcalcifications in a power law background for a free search task; the second study simu‐lates smooth or irregular microcalcifications in the same type of backgrounds for a classification task. Results: Including the resolution model in the iterative reconstruction methods increases the CNR of microcalcifications. The first observer study shows a significant improvement in detection of microcalcifications ( p = 0.029), while the second study shows that performance on a classification task remains the same ( p = 0.935) compared to the filtered backprojection method. Conclusions: The new method shows higher CNR and improved visualization of microcalcifications in an observer experiment on synthetic data. Further study of the negative results of the classification task showed performance variations throughout the volume linked to the changing noise structure introduced by the combination of the resolution model and the smoothing prior.

We present a survey of techniques for the reduction of streaking artefacts caused by metallic objects in X-ray Computed Tomography (CT) images. A comprehensive review of the existing state-of-the-art Metal Artefact Reduction (MAR) techniques, drawn predominantly from the medical CT literature, is supported by an experimental comparison of twelve MAR techniques. The experimentation is grounded in an evaluation based on a standard scientific comparison protocol for MAR methods, using a software generated medical phantom image as well as a clinical CT scan. The experimentation is extended by considering novel applications of CT imagery consisting of metal objects in non-tissue surroundings acquired from the aviation security screening domain. We address the shortage of thorough performance analyses in the existing MAR literature by conducting a qualitative as well as quantitative comparative evaluation of the selected techniques. We find that the difficulty in generating accurate priors to be the predominant factor limiting the effectiveness of the state-of-the-art medical MAR techniques when applied to non-medical CT imagery. This study thus extends previous works by: comparing several state-of-the-art MAR techniques; considering both medical and non-medical applications and performing a thorough performance analysis, considering both image quality as well as computational demands.

This paper presents an extension to a recent intensity-limiting sino-gram completion-based Metal Artefact Reduction (MAR) algorithm for Computed Tomography (CT) images containing multiple metal objects. A novel weighting scheme is introduced, whereby the intensities of the MAR-corrected pixels are modified based on their spatial locations relative to the metal objects. Pixels falling within the straight-line regions connecting multiple metal objects are subjected to less intensive intensity-limiting, thereby compensating for the characteristic dark bands occurring in these regions. Extensive experimentation is performed on a state-of-the-art numerical simulation, a clinical CT data set and a baggage security CT data set. Comprehensive performance analysis, using reference and reference-free error metrics, Bland-Altman plots and visual comparisons, demonstrate an improvement in the restoration of the underestimated intensities occurring in the regions connecting multiple metal objects.

Identification of myocardial infarction (MI) by imaging is critical for clinical management of ischemic heart disease. Iodine-123-labeled hypericin (I-123-Hyp) is a new potent infarct avid agent. We sought to compare target selectivity and organ distribution between I-123-Hyp and the myocardial perfusion agent, technetium-99m-labeled hexakis [2-methoxy isobutyl isonitrile] (Tc-99m-Sestamibi) in rabbits with acute MI. Hypericin was radiolabeled with I-123 using iodogen as oxidant, and Tc-99m-Sestamibi was prepared from a commercial kit and radioactive sodium pertechnetate. Rabbits (n = 6) with 24-hour-old MI received I-123-Hyp intravenously and received Tc-99m-Sestamibi 9 hours later. They were studied by dual-isotope simultaneous acquisition micro single photon emission computed tomography/computed tomography (DISA-SPECT/CT), tissue gamma counting (TGC), autoradiography, and histology. After purification, I-123-Hyp was obtained with radiochemical purity around 99%. DISA-SPECT/CT images showed I-123-Hyp retention in infarcted but not in normal myocardium. By TGC, accumulation values reached 1.175 +/- 0.096 percentage of injected dose per gram (%ID/g) and 0.028 +/- 0.007%ID/g in infarcted myocardium and normal myocardium with high tracer concentration in liver, intestines, and gallbladder. Tc-99m-Sestamibi was prepared with radiochemical purity over 95%. DISA-SPECT/CT showed no accumulation in MI and high initial radioactivity levels in normal myocardium that were rapidly cleared as confirmed by TGC (0.011 +/- 0.003%ID/g). Liver and intestines were clearly visualized. By TGC, gallbladder and kidneys show moderate Tc-99m-Sestamibi uptake. The selectivity of I-123-Hyp for infarcted myocardium and Tc-99m-Sestamibi for normal myocardium was confirmed. I-123-Hyp distribution in rabbits is characterized by hepatobiliary excretion. Tc-99m-Sestamibi undergoes hepatorenal elimination.

PURPOSE In iterative reconstruction, metal artifacts can be reduced by applying more accurate reconstruction models that are usually also more computationally demanding. The hypothesis of this work is that these complex models only need to be applied in the vicinity of the metals and that a less complex model can be used for the remainder of the reconstruction volume. METHODS A method is described that automatically divides the reconstruction volume into metal and nonmetal regions. The different regions are called patches. A different energy and resolution model can be assigned to each of the patches. The patches containing metals are reconstructed with a fully polychromatic spectral model (IMPACT) and if necessary with an increased resolution model. The patch without metals is reconstructed with a simple polychromatic model (MLTRC) that only includes the spectral behavior of water attenuation. Comparing the computational complexity of IMPACT and MLTRC gives a ratio of 8:3. The different patches are updated sequentially as in a grouped coordinate algorithm. Two phantoms were simulated and measured: a circular phantom containing small metal cylinders and a body phantom representing a human pelvis with two femoral implants. As a first test, the sequential update of the patches was applied while using the same energy model for all patches. Second, the local model approach was applied using MLTRC for nonmetal regions and IMPACT for metal regions. The results of different iterative reconstruction schemes are compared to the results of projection completion, another important method for the reduction of metal artifacts. RESULTS Reconstruction schemes including the sequential update of the patches result in images with less streak artifacts compared to a regular reconstruction. The sequential update of each of the metal regions improves the relative convergence of the metals (edges and attenuation values) against the rest of the image, which leads to an improved artifact reduction. Using the combined IMPACT+MLTRC model results in a similar image quality as using IMPACT everywhere, while providing an important benefit regarding computational complexity. Some streak and shadow artifacts were still present, but all structures present in the phantom could be observed. Projection completion results in reconstructions with less obvious streak and shadow artifacts but tends to deform or erase structures lying close to or in between metallic structures. CONCLUSIONS Metal artifact reduction with iterative reconstruction can be achieved by using complex models only locally without losing image quality. Separately updating metal regions leads to reduced streak artifacts. Structures lying close to or in between metals are often better reconstructed, compared to projection completion results, because all available information is used.

In this paper, the Poisson distribution near zero was modified to remove this non-negativity constraint in PET.

It is known that the convergence of iterative algorithms is better when updating pixels sequentially rather than simultaneously, but sequential updating is more computationally demanding and less practical for implementation issues. We propose to divide the image in a matrix of equally sized cuboids and update them separately as groups of pixels in a grouped coordinate ascent algorithm. We apply this to transmission and emission tomography. For both applications we see a substantial increase in convergence per iteration.

The use of iterative image reconstruction algorithms with resolution modeling allows for reduced partial volume effect without noise increase. However, it is now recognized that EM-ML type algorithms are biased in very low counts images, in particular for cold regions. Alternative ML algorithms that allow for negative image voxels have been proposed to reduce the bias: NEG-ML of Nuyts et al and AB-ML of Byrne.The aim of this study is to evaluate the NEG-ML and AB-ML algorithms with respect to the EM-ML and 3DRP algorithms for human brain dynamic studies on the high resolution HRRT scanner. As the ground truth is not known, the idea is to distribute the list-mode data into several statistically independent gates of 2ms duration and to compare the summation of the reconstructions of each gate to the reconstruction of the entire list-mode. As both images are produced using the same data, differences are solely due to low count statistics effects. The average number of events per replicate is selected by choosing the number of gates.A one hour FDG brain PET study was split into different numbers of gates ( ranging from 2 to 360). The count statistics of each replica ranged from a 30min to a 10s acquisition, resulting in weakly to extremely noisy data. Two threshold values in the denominators of the NEG-ML equation were used: 1 as originally recommended and 10(-4) as frequently used for EM-ML. Arbitrary high bound values were used for the AB-ML algorithm. The mean activity concentration was measured in the gray and white matters, based on regions of interest drawn on a MRI of the same subject. As expected, no differences were found for 3DRP between the one hour acquisition and the sum of shorter acquisitions. EM-ML and NEG-ML with the lower threshold lead to an absolute bias ranging from 0.1 to 4.5% for 30min to 10s acquisitions. The bias was reduced to less than 0.5% for AB-ML and NEG-ML with a threshold of 1, for all acquisition durations. Convergence and noise properties were also studied.The AB-ML and NEG-ML algorithms are almost bias-free alternatives to EM-ML PET reconstruction of data with any noise level. The choice of the denominator threshold of the NEG-ML algorithm seems to play an important role.

Metal artefact reduction (MAR) in computed tomography remains a challenging problem. Projection completion (PC) and iterative reconstruction with an advanced projection model are the two most important MAR methods. PC often results in strong artefact reduction, but by the interpolation step information about the metal and its surrounding is lost. This information can be important in e.g. orthopaedic surgery for implant follow-up. Iterative methods use all available information and are more suitable in such cases. Unfortunately, these methods are slow, especially when using a more adequate, more complex model. We present a local model reconstruction scheme which uses iterative reconstruction but only increases the complexity of the model in the vicinity of metals. Hereby we can limit the computation time while keeping a similar image quality. Moreover, when using a sequential update of several image parts, one obtains a better convergence of the metals, leading to an improved artefact reduction.

Artifacts in computed tomography (CT) reconstruction are often caused by an inaccurate modeling of the acquisition during reconstruction. These artifacts can be severe when metals are in the field of view. A better modeling of the acquisition during reconstruction reduces artifacts but often leads to a substantial increase in computation time. Because metals are the most prominent source of the artifacts, we hypothesize it would be sufficient to limit the more complex model to regions close to metals and use less complex, faster models for the other parts. For this purpose we present a patchwork (back) projector, embedded in an iterative reconstruction algorithm, which is able to change reconstruction properties for a particular area in the object. We combined reconstruction models with different complexity for the reduction of non-linear partial volume effects and beam hardening. In this way we reduced the computation time while keeping the same image quality. Moreover, an improved convergence is achieved by dividing the image into subareas, patches. The presented results are for a two dimensional geometry, in the future we will extend this to three dimensions.