Limited-Angle Computed Tomography (LACT) is a nondestructive 3D imaging technique used in a variety of applications ranging from security to medicine. The limited angle coverage in LACT is often a dominant source of severe artifacts in the reconstructed images, making it a challenging imaging inverse problem. Diffusion models are a recent class of deep generative models for synthesizing realistic images using image denoisers. In this work, we present DOLCE as the first framework for integrating conditionally-trained diffusion models and explicit physical measurement models for solving imaging inverse problems. DOLCE achieves the SOTA performance in highly ill-posed LACT by alternating between the data-fidelity and sampling updates of a diffusion model conditioned on the transformed sinogram. We show through extensive experimentation that unlike existing methods, DOLCE can synthesize high-quality and structurally coherent 3D volumes by using only 2D conditionally pre-trained diffusion models. We further show on several challenging real LACT datasets that the same pretrained DOLCE model achieves the SOTA performance on drastically different types of images.

As computational tools for X-ray computed tomography (CT) become more quantitatively accurate, knowledge of the source-detector spectral response is critical for quantitative system-independent reconstruction and material characterization capabilities. Directly measuring the spectral response of a CT system is hard, which motivates spectral estimation using transmission data obtained from a collection of known homogeneous objects. However, the associated inverse problem is ill-conditioned, making accurate estimation of the spectrum challenging, particularly in the absence of a close initial guess.In this paper, we describe a dictionary-based spectral estimation method that yields accurate results without the need for any initial estimate of the spectral response. Our method utilizes a MAP estimation framework that combines a physics-based forward model along with an L0sparsity constraint and a simplex constraint on the dictionary coefficients. Our method uses a greedy support selection method and a new pair-wise iterated coordinate descent method to compute the above estimate. We demonstrate that our dictionary-based method outperforms a state-of-the-art method as shown in a cross-validation experiment on four real datasets collected at beamline 8.3.2 of the Advanced Light Source (ALS).

Prognoses of Traumatic Brain Injury (TBI) outcomes are neither easily nor accurately determined from clinical indicators. This is due in part to the heterogeneity of damage inflicted to the brain, ultimately resulting in diverse and complex outcomes. Using a data-driven approach on many distinct data elements may be necessary to describe this large set of outcomes and thereby robustly depict the nuanced differences among TBI patients' recovery. In this work, we develop a method for modeling large heterogeneous data types relevant to TBI. Our approach is geared toward the probabilistic representation of mixed continuous and discrete variables with missing values. The model is trained on a dataset encompassing a variety of data types, including demographics, blood-based biomarkers, and imaging findings. In addition, it includes a set of clinical outcome assessments at 3, 6, and 12 months post-injury. The model is used to stratify patients into distinct groups in an unsupervised learning setting. We use the model to infer outcomes using input data, and show that the collection of input data reduces uncertainty of outcomes over a baseline approach. In addition, we quantify the performance of a likelihood scoring technique that can be used to self-evaluate the extrapolation risk of prognosis on unseen patients.

We present a novel neural network architecture called AutoAtlas for fully unsupervised partitioning and representation learning of 3D brain Magnetic Resonance Imaging (MRI) volumes. AutoAtlas consists of two neural network components: one neural network to perform multi-label partitioning based on local texture in the volume, and a second neural network to compress the information contained within each partition. We train both of these components simultaneously by optimizing a loss function that is designed to promote accurate reconstruction of each partition, while encouraging spatially smooth and contiguous partitioning, and discouraging relatively small partitions. We show that the partitions adapt to the subject specific structural variations of brain tissue while consistently appearing at similar spatial locations across subjects. AutoAtlas also produces very low dimensional features that represent local texture of each partition. We demonstrate prediction of metadata associated with each subject using the derived feature representations and compare the results to prediction using features derived from FreeSurfer anatomical parcellation. Since our features are intrinsically linked to distinct partitions, we can then map values of interest, such as partition-specific feature importance scores onto the brain for visualization.

Reconstructing dynamic, time-varying scenes with computed tomography (4D-CT) is a challenging and ill-posed problem common to industrial and medical settings. Existing 4D-CT reconstructions are designed for sparse sampling schemes that require fast CT scanners to capture multiple, rapid revolutions around the scene in order to generate high quality results. However, if the scene is moving too fast, then the sampling occurs along a limited view and is difficult to reconstruct due to spatiotemporal ambiguities. In this work, we design a reconstruction pipeline using implicit neural representations coupled with a novel parametric motion field warping to perform limited view 4D-CT reconstruction of rapidly deforming scenes. Importantly, we utilize a differentiable analysis-bysynthesis approach to compare with captured x-ray sinogram data in a self-supervised fashion. Thus, our resulting optimization method requires no training data to reconstruct the scene. We demonstrate that our proposed system robustly reconstructs scenes containing deformable and periodic motion and validate against state-of-the-art baselines. Further, we demonstrate an ability to reconstruct continuous spatiotemporal representations of our scenes and upsample them to arbitrary volumes and frame rates post-optimization. This research opens a new avenue for implicit neural representations in computed tomography reconstruction in general. Code is available at https://github.com/awreed/DynamicCTReconstruction.

Blur in X-ray radiographs not only reduces the sharpness of image edges but also reduces the overall contrast. The effective blur in a radiograph is the combined effect of blur from multiple sources such as the detector panel, X-ray source spot, and system motion. In this paper, we use a systems approach to model the point spread function (PSF) of the effective radiographic blur as the convolution of multiple PSFs, where each PSF models one of the various sources of blur. In particular, we model the combined contribution of X-ray source and detector blurs while assuming negligible contribution from other forms of blur. Then, we present a numerical optimization algorithm for estimating the source and detector PSFs from multiple radiographs acquired at different X-ray source to object (SOD) and object to detector distances (ODD). Finally, we computationally reduce blur in radiographs using deblurring algorithms that use the estimated PSFs from the previous step. Our approach to estimate and reduce blur is called SABER, which is an acronym for systems approach to blur estimation and reduction.

Computed Tomography (CT) reconstruction is a fundamental component to a wide variety of applications ranging from security, to healthcare. The classical techniques require measuring projections, called sinograms, from a full 180° view of the object. However, obtaining a full-view is not always feasible, such as when scanning irregular objects that limit flexibility of scanner rotation. The resulting limited angle sinograms are known to produce highly artifact-laden reconstructions with existing techniques. In this paper, we propose to address this problem using CTNet - a system of 1D and 2D convolutional neural networks, that operates directly on a limited angle sinogram to predict the reconstruction. We use the x-ray transform on this prediction to obtain a "completed" sinogram, as if it came from a full 180°view. We feed this to standard analytical and iterative reconstruction techniques to obtain the final reconstruction. We show with extensive experimentation on a challenging real world dataset that this combined strategy outperforms many competitive baselines. We also propose a measure of confidence for the reconstruction that enables a practitioner to gauge the reliability of a prediction made by CTNet. We show that this measure is a strong indicator of quality as measured by the PSNR, while not requiring ground truth at test time. Finally, using a segmentation experiment, we show that our reconstruction also preserves the 3D structure of objects better than existing solutions.

Model based iterative reconstruction (MBIR) algorithms have been used to greatly improve image quality and temporal resolution in synchrotron based time-space Computed Tomography (CT). Among the various optimization methods that have been used for MBIR, iterative coordinate descent (ICD) has relatively lower computational requirements because it converges fast. In spite of that, long execution time remains a barrier for the widespread use of MBIR. In this paper, we present a new data structure, called VL-Buffer, for time-space reconstruction that significantly improves the cache locality while retaining good parallel performance. Experimental results show an average speedup of 40% using VL-Buffer.

X-ray propagation-based phase contrast tomography (XPCT) has become popular as a method for tomographically reconstructing the complex refractive index of a material from single distance measurements. XPCT has two major advantages over traditional X-ray computed tomography (CT): It can be used at higher cone-beam magnifications, and it typically produces higher contrast. However, current XPCT reconstruction algorithms are limited to near-field diffraction, which limits both their use and the quality of reconstructions. In this paper, we present a model-based iterative reconstruction (MBIR) algorithm called complex refractive index tomographic iterative reconstruction (CRITIR). CRITIR is based on a non-linear physics based model for X-ray propagation and a prior model for the complex refractive index of the object being imaged. Unlike conventional methods, CRITIR is designed to work within and beyond the near-field diffraction region. We use simulation to show that our algorithm accurately reconstructs the object while the conventional methods result in inaccurate reconstructions with blurry edges beyond the near-field region.

X-ray tomography is increasingly being used for 4D spatio-temporal imaging of material samples at micron and finer scales. However, the temporal resolution of widely used 4D reconstruction methods is severely limited by the need to acquire a very large number of views for each reconstructed 3D volume. In this paper, we present a time interlaced model-based iterative reconstruction (TIMBIR) method which can significantly improve the temporal resolution of reconstructions. TIMBIR is a synergistic combination of two innovations. The first innovation, interlaced view sampling, is a novel approach to data acquisition which distributes the view angles more evenly in time. The second innovation is a 4D model based iterative reconstruction algorithm (MBIR) which can produce time resolved volumetric reconstructions of the sample from the interlaced views. Reconstructions of simulated data indicate that TIMBIR can improve the temporal resolution by an order of magnitude relative to existing approaches.

Synchrotron based X-ray tomography is widely used for three dimensional imaging of materials at the micron scale. Tomographic data collected from a synchrotron is often affected by non-idealities in the measurement system and sudden “blinding” of detector pixels during the acquisition. Typically, reconstructions are done using analytical reconstruction techniques combined with pre/post-processing steps to correct for the non-idealities, resulting in loss of detail while still producing noisy reconstructions with some artifacts. In this paper, we present a model-based iterative reconstruction (MBIR) algorithm for synchrotron X-ray tomography that can automatically handle the non-idealities as a part of the reconstruction. First, we develop a forward model that accounts for the non-idealities in the measurement system and for the occurrence of outliers in the measurement. Next, we combine the forward model with a prior model of the object to formulate the MBIR cost function and propose an algorithm to minimize the cost. Results on a real data set show that the MBIR reconstructions are superior to the analytical reconstructions effectively suppressing noise as well as other artifacts.