In this paper, we propose a novel deep convolutional neural network (CNN)-based algorithm for solving ill-posed inverse problems. Regularized iterative algorithms have emerged as the standard approach to ill-posed inverse problems in the past few decades. These methods produce excellent results, but can be challenging to deploy in practice due to factors including the high computational cost of the forward and adjoint operators and the difficulty of hyperparameter selection. The starting point of this paper is the observation that unrolled iterative methods have the form of a CNN (filtering followed by pointwise non-linearity) when the normal operator ( $H^{*}H$ , where $H^{*}$ is the adjoint of the forward imaging operator, $H$ ) of the forward model is a convolution. Based on this observation, we propose using direct inversion followed by a CNN to solve normal-convolutional inverse problems. The direct inversion encapsulates the physical model of the system, but leads to artifacts when the problem is ill posed; the CNN combines multiresolution decomposition and residual learning in order to learn to remove these artifacts while preserving image structure. We demonstrate the performance of the proposed network in sparse-view reconstruction (down to 50 views) on parallel beam X-ray computed tomography in synthetic phantoms as well as in real experimental sinograms. The proposed network outperforms total variation-regularized iterative reconstruction for the more realistic phantoms and requires less than a second to reconstruct a $512\times 512$ image on the GPU.

/pdf/deep-convolutional-neural-network-for-inverse-problems-in-20ejy5u0ob.pdf

Deep Convolutional Neural Network for Inverse Problems in Imaging

Given the potential risk of X-ray radiation to the patient, low-dose CT has attracted a considerable interest in the medical imaging field. Currently, the main stream low-dose CT methods include vendor-specific sinogram domain filtration and iterative reconstruction algorithms, but they need to access raw data, whose formats are not transparent to most users. Due to the difficulty of modeling the statistical characteristics in the image domain, the existing methods for directly processing reconstructed images cannot eliminate image noise very well while keeping structural details. Inspired by the idea of deep learning, here we combine the autoencoder, deconvolution network, and shortcut connections into the residual encoder–decoder convolutional neural network (RED-CNN) for low-dose CT imaging. After patch-based training, the proposed RED-CNN achieves a competitive performance relative to the-state-of-art methods in both simulated and clinical cases. Especially, our method has been favorably evaluated in terms of noise suppression, structural preservation, and lesion detection.

/pdf/low-dose-ct-with-a-residual-encoder-decoder-convolutional-2mk77wlfg8.pdf

Low-Dose CT With a Residual Encoder-Decoder Convolutional Neural Network

X-ray computer tomography (CT) is fast becoming an accepted tool within the materials science community for the acquisition of 3D images. Here the authors review the current state of the art as CT transforms from a qualitative diagnostic tool to a quantitative one. Our review considers first the image acquisition process, including the use of iterative reconstruction strategies suited to specific segmentation tasks and emerging methods that provide more insight (e.g. fast and high resolution imaging, crystallite (grain) imaging) than conventional attenuation based tomography. Methods and shortcomings of CT are examined for the quantification of 3D volumetric data to extract key topological parameters such as phase fractions, phase contiguity, and damage levels as well as density variations. As a non-destructive technique, CT is an ideal means of following structural development over time via time lapse sequences of 3D images (sometimes called 3D movies or 4D imaging). This includes information nee...

https://www.tandfonline.com/doi/pdf/10.1179/1743280413Y.0000000023

Quantitative X-ray tomography

The continuous development and extensive use of computed tomography (CT) in medical practice has raised a public concern over the associated radiation dose to the patient. Reducing the radiation dose may lead to increased noise and artifacts, which can adversely affect the radiologists’ judgment and confidence. Hence, advanced image reconstruction from low-dose CT data is needed to improve the diagnostic performance, which is a challenging problem due to its ill-posed nature. Over the past years, various low-dose CT methods have produced impressive results. However, most of the algorithms developed for this application, including the recently popularized deep learning techniques, aim for minimizing the mean-squared error (MSE) between a denoised CT image and the ground truth under generic penalties. Although the peak signal-to-noise ratio is improved, MSE- or weighted-MSE-based methods can compromise the visibility of important structural details after aggressive denoising. This paper introduces a new CT image denoising method based on the generative adversarial network (GAN) with Wasserstein distance and perceptual similarity. The Wasserstein distance is a key concept of the optimal transport theory and promises to improve the performance of GAN. The perceptual loss suppresses noise by comparing the perceptual features of a denoised output against those of the ground truth in an established feature space, while the GAN focuses more on migrating the data noise distribution from strong to weak statistically. Therefore, our proposed method transfers our knowledge of visual perception to the image denoising task and is capable of not only reducing the image noise level but also trying to keep the critical information at the same time. Promising results have been obtained in our experiments with clinical CT images.

Low-Dose CT Image Denoising Using a Generative Adversarial Network With Wasserstein Distance and Perceptual Loss

Model-based reconstruction is a powerful framework for solving a variety of inverse problems in imaging. In recent years, enormous progress has been made in the problem of denoising, a special case of an inverse problem where the forward model is an identity operator. Similarly, great progress has been made in improving model-based inversion when the forward model corresponds to complex physical measurements in applications such as X-ray CT, electron-microscopy, MRI, and ultrasound, to name just a few. However, combining state-of-the-art denoising algorithms (i.e., prior models) with state-of-the-art inversion methods (i.e., forward models) has been a challenge for many reasons. In this paper, we propose a flexible framework that allows state-of-the-art forward models of imaging systems to be matched with state-of-the-art priors or denoising models. This framework, which we term as Plug-and-Play priors, has the advantage that it dramatically simplifies software integration, and moreover, it allows state-of-the-art denoising methods that have no known formulation as an optimization problem to be used. We demonstrate with some simple examples how Plug-and-Play priors can be used to mix and match a wide variety of existing denoising models with a tomographic forward model, thus greatly expanding the range of possible problem solutions.

/pdf/plug-and-play-priors-for-model-based-reconstruction-51xolwckvh.pdf

Plug-and-Play priors for model based reconstruction

Although diagnostic medical imaging provides enormous benefits in the early detection and accuracy diagnosis of various diseases, there are growing concerns on the potential side effect of radiation induced genetic, cancerous and other diseases. How to reduce radiation dose while maintaining the diagnostic performance is a major challenge in the computed tomography (CT) field. Inspired by the compressive sensing theory, the sparse constraint in terms of total variation (TV) minimization has already led to promising results for low-dose CT reconstruction. Compared to the discrete gradient transform used in the TV method, dictionary learning is proven to be an effective way for sparse representation. On the other hand, it is important to consider the statistical property of projection data in the low-dose CT case. Recently, we have developed a dictionary learning based approach for low-dose X-ray CT. In this paper, we present this method in detail and evaluate it in experiments. In our method, the sparse constraint in terms of a redundant dictionary is incorporated into an objective function in a statistical iterative reconstruction framework. The dictionary can be either predetermined before an image reconstruction task or adaptively defined during the reconstruction process. An alternating minimization scheme is developed to minimize the objective function. Our approach is evaluated with low-dose X-ray projections collected in animal and human CT studies, and the improvement associated with dictionary learning is quantified relative to filtered backprojection and TV-based reconstructions. The results show that the proposed approach might produce better images with lower noise and more detailed structural features in our selected cases. However, there is no proof that this is true for all kinds of structures.

/pdf/low-dose-x-ray-ct-reconstruction-via-dictionary-learning-3jx69t6xok.pdf

Low-Dose X-ray CT Reconstruction via Dictionary Learning

X-ray micro-CT is an important imaging tool for biomedical researchers Our group has recently proposed a hybrid “true-color” micro-CT system to improve contrast resolution with lower system cost and radiation dose The system incorporates an energy-resolved photon-counting true-color detector into a conventional micro-CT configuration, and can be used for material decomposition In this paper, we demonstrate an interior color-CT image reconstruction algorithm developed for this hybrid true-color micro-CT system A compressive sensing-based statistical interior tomography method is employed to reconstruct each channel in the local spectral imaging chain, where the reconstructed global gray-scale image from the conventional imaging chain served as the initial guess Principal component analysis was used to map the spectral reconstructions into the color space The proposed algorithm was evaluated by numerical simulations, physical phantom experiments, and animal studies The results confirm the merits of the proposed algorithm, and demonstrate the feasibility of the hybrid true-color micro-CT system Additionally, a ``color diffusion'' phenomenon was observed whereby high-quality true-color images are produced not only inside the region of interest, but also in neighboring regions It appears harnessing that this phenomenon could potentially reduce the color detector size for a given ROI, further reducing system cost and radiation dose

Image Reconstruction for Hybrid True-Color Micro-CT

This paper presents a statistical interior tomography (SIT) approach making use of compressed sensing (CS) theory. With the projection data modeled by the Poisson distribution, an objective function with a total variation (TV) regularization term is formulated in the maximization of a posteriori (MAP) framework to solve the interior problem. An alternating minimization method is used to optimize the objective function with an initial image from the direct inversion of the truncated Hilbert transform. The proposed SIT approach is extensively evaluated with both numerical and real datasets. The results demonstrate that SIT is robust with respect to data noise and down-sampling, and has better resolution and less bias than its deterministic counterpart in the case of low count data.

/pdf/statistical-interior-tomography-rcll8h2rrx.pdf

Statistical Interior Tomography

Scatter correction is an open problem in x-ray cone beam (CB) CT. The measurement of scatter intensity with a moving beam stop array (BSA) is a promising technique that offers a low patient dose and accurate scatter measurement. However, when restoring the blocked primary fluence behind the BSA, spatial interpolation cannot well restore the high-frequency part, causing streaks in the reconstructed image. To address this problem, we deduce a projection correlation (PC) to utilize the redundancy (over-determined information) in neighbouring CB views. PC indicates that the main high-frequency information is contained in neighbouring angular projections, instead of the current projection itself, which provides a guiding principle that applies to high-frequency information restoration. On this basis, we present the projection correlation based view interpolation (PC-VI) algorithm; that it outperforms the use of only spatial interpolation is validated. The PC-VI based moving BSA method is developed. In this method, PC-VI is employed instead of spatial interpolation, and new moving modes are designed, which greatly improve the performance of the moving BSA method in terms of reliability and practicability. Evaluation is made on a high-resolution voxel-based human phantom realistically including the entire procedure of scatter measurement with a moving BSA, which is simulated by analytical ray-tracing plus Monte Carlo simulation with EGSnrc. With the proposed method, we get visually artefact-free images approaching the ideal correction. Compared with the spatial interpolation based method, the relative mean square error is reduced by a factor of 6.05-15.94 for different slices. PC-VI does well in CB redundancy mining; therefore, it has further potential in CBCT studies.

http://iopscience.iop.org/article/10.1088/0031-9155/55/21/002/pdf

Projection correlation based view interpolation for cone beam CT: primary fluence restoration in scatter measurement with a moving beam stop array.

The long-standing interior problem has been recently revisited, leading to promising results on exact local reconstruction
also referred to as interior tomography. To date, there are two key computational ingredients of interior tomography. The
first ingredient is inversion of the truncated Hilbert transform with prior sub-region knowledge. The second is
compressed sensing (CS) assuming a piecewise constant or polynomial region of interest (ROI). Here we propose a
statistical approach for interior tomography incorporating the aforementioned two ingredients as well. In our approach,
projection data follows the Poisson model, and an image is reconstructed in the maximum a posterior (MAP) framework
subject to other interior tomography constraints including known subregion and minimized total variation (TV). A
deterministic interior reconstruction based on the inversion of the truncated Hilbert transform is used as the initial image
for the statistical interior reconstruction. This algorithm has been extensively evaluated in numerical and animal studies
in terms of major image quality indices, radiation dose and machine time. In particular, our encouraging results from a
low-contrast Shepp-Logan phantom and a real sheep scan demonstrate the feasibility and merits of our proposed
statistical interior tomography approach.

Qiong Xu

Papers

Low-Dose X-ray CT Reconstruction via Dictionary Learning

Image Reconstruction for Hybrid True-Color Micro-CT

Statistical Interior Tomography

Projection correlation based view interpolation for cone beam CT: primary fluence restoration in scatter measurement with a moving beam stop array.

Statistical interior tomography