A denoising diffusion probabilistic model for metal artifact reduction in CT.

The presence of metal objects leads to corrupted CT projection measurements, resulting in metal artifacts in the reconstructed CT images. AI promises to offer improved solutions to estimate missing sinogram data for metal artifact reduction (MAR), as previously shown with convolutional neural networks (CNNs) and generative adversarial networks (GANs). Recently, denoising diffusion probabilistic models (DDPM) have shown great promise in image generation tasks, potentially outperforming GANs. In this study, a DDPM-based approach is proposed for inpainting of missing sinogram data for improved MAR. The proposed model is unconditionally trained, free from information on metal objects, which can potentially enhance its generalization capabilities across different types of metal implants compared to conditionally trained approaches. The performance of the proposed technique was evaluated and compared to the state-of-the-art normalized MAR (NMAR) approach as well as to CNN-based and GAN-based MAR approaches. The DDPM-based approach provided significantly higher SSIM and PSNR, as compared to NMAR (SSIM: p < 10-26; PSNR: p < 10-21), the CNN (SSIM: p < 10-25; PSNR: p < 10-9) and the GAN (SSIM: p < 10-6; PSNR: p < 0.05) methods. The DDPM-MAR technique was further evaluated based on clinically relevant image quality metrics on clinical CT images with virtually introduced metal objects and metal artifacts, demonstrating superior quality relative to the other three models. In general, the AI-based techniques showed improved MAR performance compared to the non-AI-based NMAR approach. The proposed methodology shows promise in enhancing the effectiveness of MAR, and therefore improving the diagnostic accuracy of CT.

Simulated deep CT characterization of liver metastases with high-resolution filtered back projection reconstruction.

Early diagnosis and accurate prognosis of colorectal cancer is critical for determining optimal treatment plans and maximizing patient outcomes, especially as the disease progresses into liver metastases. Computed tomography (CT) is a frontline tool for this task; however, the preservation of predictive radiomic features is highly dependent on the scanning protocol and reconstruction algorithm. We hypothesized that image reconstruction with a high-frequency kernel could result in a better characterization of liver metastases features via deep neural networks. This kernel produces images that appear noisier but preserve more sinogram information. A simulation pipeline was developed to study the effects of imaging parameters on the ability to characterize the features of liver metastases. This pipeline utilizes a fractal approach to generate a diverse population of shapes representing virtual metastases, and then it superimposes them on a realistic CT liver region to perform a virtual CT scan using CatSim. Datasets of 10,000 liver metastases were generated, scanned, and reconstructed using either standard or high-frequency kernels. These data were used to train and validate deep neural networks to recover crafted metastases characteristics, such as internal heterogeneity, edge sharpness, and edge fractal dimension. In the absence of noise, models scored, on average, 12.2% ( α = 0.012 ) and 7.5% ( α = 0.049 ) lower squared error for characterizing edge sharpness and fractal dimension, respectively, when using high-frequency reconstructions compared to standard. However, the differences in performance were statistically insignificant when a typical level of CT noise was simulated in the clinical scan. Our results suggest that high-frequency reconstruction kernels can better preserve information for downstream artificial intelligence-based radiomic characterization, provided that noise is limited. Future work should investigate the information-preserving kernels in datasets with clinical labels.

Deep-silicon photon-counting x-ray projection denoising through reinforcement learning.

BACKGROUND: In recent years, deep reinforcement learning (RL) has been applied to various medical tasks and produced encouraging results. OBJECTIVE: In this paper, we demonstrate the feasibility of deep RL for denoising simulated deep-silicon photon-counting CT (PCCT) data in both full and interior scan modes. PCCT offers higher spatial and spectral resolution than conventional CT, requiring advanced denoising methods to suppress noise increase. METHODS: In this work, we apply a dueling double deep Q network (DDDQN) to denoise PCCT data for maximum contrast-to-noise ratio (CNR) and a multi-agent approach to handle data non-stationarity. RESULTS: Using our method, we obtained significant image quality improvement for single-channel scans and consistent improvement for all three channels of multichannel scans. For the single-channel interior scans, the PSNR (dB) and SSIM increased from 33.4078 and 0.9165 to 37.4167 and 0.9790 respectively. For the multichannel interior scans, the channel-wise PSNR (dB) increased from 31.2348, 30.7114, and 30.4667 to 31.6182, 30.9783, and 30.8427 respectively. Similarly, the SSIM improved from 0.9415, 0.9445, and 0.9336 to 0.9504, 0.9493, and 0.0326 respectively. CONCLUSIONS: Our results show that the RL approach improves image quality effectively, efficiently, and consistently across multiple spectral channels and has great potential in clinical applications.

Development and tuning of models for accurate simulation of CT spatial resolution using CatSim.

Objective. We sought to systematically evaluate CatSim's ability to accurately simulate the spatial resolution produced by a typical 64-detector-row clinical CT scanner in the projection and image domains, over the range of clinically used x-ray techniques.Approach.Using a 64-detector-row clinical scanner, we scanned two phantoms designed to evaluate spatial resolution in the projection and image domains. These empirical scans were performed over the standard clinically used range of x-ray techniques (kV, and mA). We extracted projection data from the scanner, and we reconstructed images. For the CatSim simulations, we developed digital phantoms to represent the phantoms used in the empirical scans. We developed a new, realistic model for the x-ray source focal spot, and we empirically tuned a published model for the x-ray detector temporal response. We applied these phantoms and models to simulate scans equivalent to the empirical scans, and we reconstructed the simulated projections using the same methods used for the empirical scans. For the empirical and simulated scans, we qualitatively and quantitatively compared the projection-domain and image-domain point-spread functions (PSFs) as well as the image-domain modulation transfer functions. We reported four quantitative metrics and the percent error between the empirical and simulated results.Main Results.Qualitatively, the PSFs matched well in both the projection and image domains. Quantitatively, all four metrics generally agreed well, with most of the average errors substantially less than 5% for all x-ray techniques. Although the errors tended to increase with decreasing kV, we found that the CatSim simulations agreed with the empirical scans within limits required for the anticipated applications of CatSim.Significance.The new focal spot model and the new detector temporal response model are significant contributions to CatSim because they enabled achieving the desired level of agreement between empirical and simulated results. With these new models and this validation, CatSim users can be confident that the spatial resolution represented by simulations faithfully represents results that would be obtained by a real scanner, within reasonable, known limits. Furthermore, users of CatSim can vary parameters including but not limited to system geometry, focal spot size/shape and detector parameters, beyond the values available in physical scanners, and be confident in the results. Therefore, CatSim can be used to explore new hardware designs as well as new scanning and reconstruction methods, thus enabling acceleration of improved CT scan capabilities.

Cardiac CT blooming artifacts: clinical significance, root causes and potential solutions.

This review paper aims to summarize cardiac CT blooming artifacts, how they present clinically and what their root causes and potential solutions are. A literature survey was performed covering any publications with a specific interest in calcium blooming and stent blooming in cardiac CT. The claims from literature are compared and interpreted, aiming at narrowing down the root causes and most promising solutions for blooming artifacts. More than 30 journal publications were identified with specific relevance to blooming artifacts. The main reported causes of blooming artifacts are the partial volume effect, motion artifacts and beam hardening. The proposed solutions are classified as high-resolution CT hardware, high-resolution CT reconstruction, subtraction techniques and post-processing techniques, with a special emphasis on deep learning (DL) techniques. The partial volume effect is the leading cause of blooming artifacts. The partial volume effect can be minimized by increasing the CT spatial resolution through higher-resolution CT hardware or advanced high-resolution CT reconstruction. In addition, DL techniques have shown great promise to correct for blooming artifacts. A combination of these techniques could avoid repeat scans for subtraction techniques.

Deep learning tomographic reconstruction through hierarchical decomposition of domain transforms.

Deep learning (DL) has shown unprecedented performance for many image analysis and image enhancement tasks. Yet, solving large-scale inverse problems like tomographic reconstruction remains challenging for DL. These problems involve non-local and space-variant integral transforms between the input and output domains, for which no efficient neural network models are readily available. A prior attempt to solve tomographic reconstruction problems with supervised learning relied on a brute-force fully connected network and only allowed reconstruction with a 1284 system matrix size. This cannot practically scale to realistic data sizes such as 5124 and 5126 for three-dimensional datasets. Here we present a novel framework to solve such problems with DL by casting the original problem as a continuum of intermediate representations between the input and output domains. The original problem is broken down into a sequence of simpler transformations that can be well mapped onto an efficient hierarchical network architecture, with exponentially fewer parameters than a fully connected network would need. We applied the approach to computed tomography (CT) image reconstruction for a 5124 system matrix size. This work introduces a new kind of data-driven DL solver for full-size CT reconstruction without relying on the structure of direct (analytical) or iterative (numerical) inversion techniques. This work presents a feasibility demonstration of full-scale learnt reconstruction, whereas more developments will be needed to demonstrate superiority relative to traditional reconstruction approaches. The proposed approach is also extendable to other imaging problems such as emission and magnetic resonance reconstruction. More broadly, hierarchical DL opens the door to a new class of solvers for general inverse problems, which could potentially lead to improved signal-to-noise ratio, spatial resolution and computational efficiency in various areas.

XCIST-an open access x-ray/CT simulation toolkit.

Objective. X-ray-based imaging modalities including mammography and computed tomography (CT) are widely used in cancer screening, diagnosis, staging, treatment planning, and therapy response monitoring. Over the past few decades, improvements to these modalities have resulted in substantially improved efficacy and efficiency, and substantially reduced radiation dose and cost. However, such improvements have evolved more slowly than would be ideal because lengthy preclinical and clinical evaluation is required. In many cases, new ideas cannot be evaluated due to the high cost of fabricating and testing prototypes. Wider availability of computer simulation tools could accelerate development of new imaging technologies. This paper introduces the development of a new open-access simulation environment for x-ray-based imaging. The main motivation of this work is to publicly distribute a fast but accurate ray-tracing x-ray and CT simulation tool along with realistic phantoms and 3D reconstruction capability, building on decades of developments in industry and academia.Approach. The x-ray-based Cancer Imaging Simulation Toolkit (XCIST) is developed in the context of cancer imaging, but can more broadly be applied. XCIST is physics-based, written in Python and C/C++, and currently consists of three major subsets: digital phantoms, the simulator itself (CatSim), and image reconstruction algorithms; planned future features include a fast dose-estimation tool and rigorous validation. To enable broad usage and to model and evaluate new technologies, XCIST is easily extendable by other researchers. To demonstrate XCIST's ability to produce realistic images and to show the benefits of using XCIST for insight into the impact of separate physics effects on image quality, we present exemplary simulations by varying contributing factors such as noise and sampling.Main results. The capabilities and flexibility of XCIST are demonstrated, showing easy applicability to specific simulation problems. Geometric and x-ray attenuation accuracy are shown, as well as XCIST's ability to model multiple scanner and protocol parameters, and to attribute fundamental image quality characteristics to specific parameters.Significance. This work represents an important first step toward the goal of creating an open-access platform for simulating existing and emerging x-ray-based imaging systems. While numerous simulation tools exist, we believe the combined XCIST toolset provides a unique advantage in terms of modeling capabilities versus ease of use and compute time. We publicly share this toolset to provide an environment for scientists to accelerate and improve the relevance of their research in x-ray and CT.

Projection decomposition via univariate optimization for dual-energy CT.

Dual-energy computed tomography (DECT) acquires two x-ray projection datasets with different x-ray energy spectra, performs material-specific image reconstruction based on the energy-dependent non-linear integral model, and provides more accurate quantification of attenuation coefficients than single energy spectrum CT. In the diagnostic energy range, x-ray energy-dependent attenuation is mainly caused by photoelectric absorption and Compton scattering. Theoretically, these two physical components of the x-ray attenuation mechanism can be determined from two projection datasets with distinct energy spectra. Practically, the solution of the non-linear integral equation is complicated due to spectral uncertainty, detector sensitivity, and data noise. Conventional multivariable optimization methods are prone to local minima. In this paper, we develop a new method for DECT image reconstruction in the projection domain. This method combines an analytic solution of a polynomial equation and a univariate optimization to solve the polychromatic non-linear integral equation. The polynomial equation of an odd order has a unique real solution with sufficient accuracy for image reconstruction, and the univariate optimization can achieve the global optimal solution, allowing accurate and stable projection decomposition for DECT. Numerical and physical phantom experiments are performed to demonstrate the effectiveness of the method in comparison with the state-of-the-art projection decomposition methods. As a result, the univariate optimization method yields a quality improvement of 15% for image reconstruction and substantial reduction of the computational time, as compared to the multivariable optimization methods.

Semiempirical, parameterized spectrum estimation for x-ray computed tomography.

PURPOSE: To develop a tool to produce accurate, well-validated x-ray spectra for standalone use or for use in an open-access x-ray/CT simulation tool. Spectrum models will be developed for tube voltages in the range of 80 kVp through 140 kVp and for anode takeoff angles in the range of 5° to 9°. METHODS: Spectra were initialized based on physics models, then refined using empirical measurements, as follows. A new spectrum-parameterization method was developed, including 13 spline knots to represent the bremsstrahlung component and 4 values to represent characteristic lines. Initial spectra at 80, 100, 120, and 140 kVp and at takeoff angles from 5° to 9° were produced using physics-based spectrum estimation tools XSPECT and SpekPy. Empirical experiments were systematically designed with careful selection of attenuator materials and thicknesses, and by reducing measurement contamination from scatter to <1%. Measurements were made on a 64-row CT scanner using the scanner's detector and using multiple layers of polymethylmethacrylate (PMMA), aluminum, titanium, tin, and neodymium. Measurements were made at 80, 100, 120, and 140 kVp and covering the entire 64-row detector (takeoff angles from 5° to 9°); a total of 6,144 unique measurements were made. After accounting for the detector's energy response, parameterized representations of the initial spectra were refined for best agreement with measurements using two proposed optimization schemes: based on modulation and based on gradient descent. X-ray transmission errors were computed for measurements vs calculations using the nonoptimized and optimized spectra. Half-value, tenth-value, and hundredth-value layers for PMMA, Al, and Ti were calculated. RESULTS: Spectra before and after parameterization were in excellent agreement (e.g., R2 values of 0.995 and 0.997). Empirical measurements produced smoothly varying curves with x-ray transmission covering a range of up to 3.5 orders of magnitude. Spectra from the two optimization schemes, compared with the unoptimized physic-based spectra, each improved agreement with measurements by twofold through tenfold, for both postlog transmission data and for fractional value layers. CONCLUSION: The resulting well-validated spectra are appropriate for use in the open-access x-ray/CT simulator under development, the x-ray-based Cancer Imaging Toolkit (XCIST), or for standalone use. These spectra can be readily interpolated to produce spectra at arbitrary kVps over the range of 80 to 140 kVp and arbitrary takeoff angles over the range of 5° to 9°. Furthermore, interpolated spectra over these ranges can be obtained by applying the standalone Matlab function available at https://github.com/xcist/documentation/blob/master/XCISTspectrum.m.

Monochromatic image reconstruction via machine learning.

X-ray computed tomography (CT) is a nondestructive imaging technique to reconstruct cross-sectional images of an object using x-ray measurements taken from different view angles for medical diagnosis, therapeutic planning, security screening, and other applications. In clinical practice, the x-ray tube emits polychromatic x-rays, and the x-ray detector array operates in the energy-integrating mode to acquire energy intensity. This physical process of x-ray imaging is accurately described by an energy-dependent non-linear integral equation on the basis of the Beer-Lambert law. However, the non-linear model is not invertible using a computationally efficient solution and is often approximated as a linear integral model in the form of the Radon transform, which basically loses energy-dependent information. This approximate model produces an inaccurate quantification of attenuation images, suffering from beam-hardening effects. In this paper, a machine learning-based approach is proposed to correct the model mismatch to achieve quantitative CT imaging. Specifically, a one-dimensional network model is proposed to learn a non-linear transform from a training dataset to map a polychromatic CT image to its monochromatic sinogram at a pre-specified energy level, realizing virtual monochromatic (VM) imaging effectively and efficiently. Our results show that the proposed method recovers high-quality monochromatic projections with an average relative error of less than 2%. The resultant x-ray VM imaging can be applied for beam-hardening correction, material differentiation and tissue characterization, and proton therapy treatment planning.

Virtual Monoenergetic CT Imaging via Deep Learning.

Conventional single-spectrum computed tomography (CT) reconstructs a spectrally integrated attenuation image and reveals tissues morphology without any information about the elemental composition of the tissues. Dual-energy CT (DECT) acquires two spectrally distinct datasets and reconstructs energy-selective (virtual monoenergetic [VM]) and material-selective (material decomposition) images. However, DECT increases system complexity and radiation dose compared with single-spectrum CT. In this paper, a deep learning approach is presented to produce VM images from single-spectrum CT images. Specifically, a modified residual neural network (ResNet) model is developed to map single-spectrum CT images to VM images at pre-specified energy levels. This network is trained on clinical DECT data and shows excellent convergence behavior and image accuracy compared with VM images produced by DECT. The trained model produces high-quality approximations of VM images with a relative error of less than 2%. This method enables multi-material decomposition into three tissue classes, with accuracy comparable with DECT.

A two-dimensional feasibility study of deep learning-based feature detection and characterization directly from CT sinograms.

Machine Learning, especially deep learning, has been used in typical x-ray computed tomography (CT) applications, including image reconstruction, image enhancement, image domain feature detection and image domain feature characterization. To our knowledge, this is the first study on machine learning for feature detection and analysis directly based on CT projection data. Specifically, we present neural network methods for blood vessel detection and characterization in the sinogram domain avoiding any partial volume, beam hardening, or motion artifacts introduced during reconstruction. First, we estimate sinogram domain vessel maps using a residual encoder-decoder convolutional neural network (REDCNN). Next, we estimate the vessel centerline and we extract the vessel-only sinogram from the original sinogram, eliminating any background information. Finally, we use a fully connected neural network to estimate the vessel lumen cross-sectional area from the vessel-only sinogram. We trained and tested the proposed methods using CatSim simulations, real CT measurements of vessel phantoms, and clinical data from the NIH CT image database. We achieved encouraging initial results showing the feasibility of CT analysis in the sinogram domain. In principle, sinogram domain analysis should be possible for many other and more complicated clinical CT analysis tasks. Further studies are needed for this sinogram domain analysis approach to become practical for clinical applications.

A dual-stream deep convolutional network for reducing metal streak artifacts in CT images.

Machine learning and deep learning are rapidly finding applications in the medical imaging field. In this paper, we address the long-standing problem of metal artifacts in computed tomography (CT) images by training a dual-stream deep convolutional neural network for streak removal. While many metal artifact reduction methods exist, even state-of-the-art algorithms fall short in some clinical applications. Specifically, proton therapy planning requires high image quality with accurate tumor volumes to ensure treatment success. We explore a dual-stream deep network structure with residual learning to correct metal streak artifacts after a first-pass by a state-of-the-art interpolation-based algorithm, NMAR. We provide the network with a mask of the streaks in order to focus attention on those areas. Our experiments compare a mean squared error loss function with a perceptual loss function to emphasize preservation of image features and texture. Both visual and quantitative metrics are used to assess the resulting image quality for metal implant cases. Success may be due to the duality of information processing, with one network stream performing local structure correction, while the other stream provides an attention mechanism to destreak effectively. This study shows that image-domain deep learning can be highly effective for metal artifact reduction (MAR), and highlights the benefits and drawbacks of different loss functions for solving a major CT reconstruction challenge.

Model-based dose reconstruction for CT dose estimation.

PURPOSE: Our goal is to develop a model-based approach for CT dose estimation. We previously presented a CT dose estimation method that offered good accuracy in soft tissue regions but lower accuracy in bone regions. In this work, we propose an improved physic-based approach to achieve high accuracy for any materials and realistic clinical anatomies. METHODS: Like Monte Carlo techniques, we start from a model or image of the patient and we model all relevant x-ray interaction processes. Unlike Monte Carlo techniques, we do not track each individual photon, but we compute the average behavior of the x-ray interactions, combining pencil-beam calculations for the first-order interactions and kernels for the higher order interactions. The new algorithm more accurately models the variation of materials in the human body, especially for higher attenuation materials such as bone, as well as the various x-ray attenuation processes. We performed validation experiments with analytic phantoms and a polychromatic x-ray spectrum, comparing to Monte Carlo simulation (GEANT4) as the ground truth. RESULTS: The results show that the proposed method has improved accuracy in both soft tissue region and bone region: less than 6% voxel-wise errors and less than 3.2% ROI-based errors in an anthropomorphic phantom. The computational cost is on the order of a low-resolution filtered backprojection reconstruction. CONCLUSIONS: We introduced improved physics-based models in a fast CT dose reconstruction approach. The improved approach demonstrated quantitatively good correspondence to a Monte Carlo gold standard in both soft tissue and bone regions in a chest phantom with a realistic polychromatic spectrum and could potentially be used for real-time applications such as patient- and organ-specific scan planning and organ dose reporting.

Dual-Energy Computed Tomography: Physical Principles, Approaches to Scanning, Usage, and Implementation: Part 1.

There are increasing applications of dual-energy computed tomography (CT), a type of spectral CT, in neuroradiology and head and neck imaging. In this 2-part review, the fundamental principles underlying spectral CT scanning and the major considerations in implementing this type of scanning in clinical practice are reviewed. In the first part of this 2-part review, the physical principles underlying spectral CT scanning are reviewed, followed by an overview of the different approaches for spectral CT scanning, including a discussion of the strengths and challenges encountered with each approach.

Dual-Energy Computed Tomography: Physical Principles, Approaches to Scanning, Usage, and Implementation: Part 2.

There are increasing applications and use of spectral computed tomography or dual-energy computed tomography (DECT) in neuroradiology and head and neck imaging in routine clinical practice. Part 1 of this 2-part review covered fundamental physical principles underlying DECT scanning and the different approaches for scanning. Part 2 focuses on important and practical considerations for implementing and using DECT in clinical practice, including a review of different images and reconstructions produced by these scanners and important and practical issues, ranging from image quality and radiation dose to workflow-related aspects of DECT scanning, that routinely come up during operationalization of DECT.

Model and reconstruction of a K-edge contrast agent distribution with an X-ray photon-counting detector.

Contrast-enhanced computed tomography (CECT) helps enhance the visibility for tumor imaging. When a high-Z contrast agent interacts with X-rays across its K-edge, X-ray photoelectric absorption would experience a sudden increment, resulting in a significant difference of the X-ray transmission intensity between the left and right energy windows of the K-edge. Using photon-counting detectors, the X-ray intensity data in the left and right windows of the K-edge can be measured simultaneously. The differential information of the two kinds of intensity data reflects the contrast-agent concentration distribution. K-edge differences between various matters allow opportunities for the identification of contrast agents in biomedical applications. In this paper, a general radon transform is established to link the contrast-agent concentration to X-ray intensity measurement data. An iterative algorithm is proposed to reconstruct a contrast-agent distribution and tissue attenuation background simultaneously. Comprehensive numerical simulations are performed to demonstrate the merits of the proposed method over the existing K-edge imaging methods. Our results show that the proposed method accurately quantifies a distribution of a contrast agent, optimizing the contrast-to-noise ratio at a high dose efficiency.

Metal artifacts in computed tomography for radiation therapy planning: dosimetric effects and impact of metal artifact reduction.

A significant and increasing number of patients receiving radiation therapy present with metal objects close to, or even within, the treatment area, resulting in artifacts in computed tomography (CT) imaging, which is the most commonly used imaging method for treatment planning in radiation therapy. In the presence of metal implants, such as dental fillings in treatment of head-and-neck tumors, spinal stabilization implants in spinal or paraspinal treatment or hip replacements in prostate cancer treatments, the extreme photon absorption by the metal object leads to prominent image artifacts. Although current CT scanners include a series of correction steps for beam hardening, scattered radiation and noisy measurements, when metal implants exist within or close to the treatment area, these corrections do not suffice. CT metal artifacts affect negatively the treatment planning of radiation therapy either by causing difficulties to delineate the target volume or by reducing the dose calculation accuracy. Various metal artifact reduction (MAR) methods have been explored in terms of improvement of organ delineation and dose calculation in radiation therapy treatment planning, depending on the type of radiation treatment and location of the metal implant and treatment site. Including a brief description of the available CT MAR methods that have been applied in radiation therapy, this article attempts to provide a comprehensive review on the dosimetric effect of the presence of CT metal artifacts in treatment planning, as reported in the literature, and the potential improvement suggested by different MAR approaches. The impact of artifacts on the treatment planning and delivery accuracy is discussed in the context of different modalities, such as photon external beam, brachytherapy and particle therapy, as well as by type and location of metal implants.

Comparison Between Pre-Log and Post-Log Statistical Models in Ultra-Low-Dose CT Reconstruction.

X-ray detectors in clinical computed tomography (CT) usually operate in current-integrating mode. Their complicated signal statistics often lead to intractable likelihood functions for practical use in model-based image reconstruction (MBIR). It is therefore desirable to design simplified statistical models without losing the essential factors. Depending on whether the CT transmission data are logarithmically transformed, pre-log and post-log models are two major categories of choices in CT MBIR. Both being approximations, it remains an open question whether one model can notably improve image quality over the other on real scanners. In this study, we develop and compare several pre-log and post-log MBIR algorithms under a unified framework. Their reconstruction accuracy based on simulation and clinical datasets are evaluated. The results show that pre-log MBIR can achieve notably better quantitative accuracy than post-log MBIR in ultra-low-dose CT, although in less extreme cases, post-log MBIR with handcrafted pre-processing remains a competitive alternative. Pre-log MBIR could play a growing role in emerging ultra-low-dose CT applications.

Direct Reconstruction of CT-based Attenuation Correction Images for PET with Cluster-Based Penalties.

Extremely low-dose CT acquisitions used for PET attenuation correction have high levels of noise and potential bias artifacts due to photon starvation. This work explores the use of a priori knowledge for iterative image reconstruction of the CT-based attenuation map. We investigate a maximum a posteriori framework with cluster-based multinomial penalty for direct iterative coordinate decent (dICD) reconstruction of the PET attenuation map. The objective function for direct iterative attenuation map reconstruction used a Poisson log-likelihood data fit term and evaluated two image penalty terms of spatial and mixture distributions. The spatial regularization is based on a quadratic penalty. For the mixture penalty, we assumed that the attenuation map may consist of four material clusters: air+background, lung, soft tissue, and bone. Using simulated noisy sinogram data, dICD reconstruction was performed with different strengths of the spatial and mixture penalties. The combined spatial and mixture penalties reduced the RMSE by roughly 2 times compared to a weighted least square and filtered backprojection reconstruction of CT images. The combined spatial and mixture penalties resulted in only slightly lower RMSE compared to a spatial quadratic penalty alone. For direct PET attenuation map reconstruction from ultra-low dose CT acquisitions, the combination of spatial and mixture penalties offers regularization of both variance and bias and is a potential method to reconstruct attenuation maps with negligible patient dose. The presented results, using a best-case histogram suggest that the mixture penalty does not offer a substantive benefit over conventional quadratic regularization and diminishes enthusiasm for exploring future application of the mixture penalty.