The efficacy of interventional treatments highly relies on an accurate identification of the target lesions and the interventional tools in the guidance images. Whereas X-ray radiography poses low doses to the patient, its weakness is in the superposition of the different image structures in a 2D image. Cone-beam computed tomography (CBCT) might look ideal providing exact 3D information, however this is at the cost of a higher radiation dose, longer imaging time, and more space requirements in the operating room. Introducing some depth information with relatively low dose, and requiring less space, digital tomosynthesis (DTS) is a potential candidate for guiding interventions. However, due to the few number of projections and to the limited angle acquisition, DTS has poor depth resolution. Since high quality patient-specific prior CT scans are usually performed prior to the intervention for diagnosis or to plan the intervention, and given that such images share a fair amount of information with the intraoperative DTS images, we propose in this work a prior-based iterative reconstruction framework to improve the intraoperative DTS image quality. The framework is based on registering the prior CT image to an intermediate low-quality intraoperative DTS image, then iteratively re-reconstructing the intraoperative DTS image using the co-registered prior CT as the starting image. We acquired prior CT and intraoperative CBCT data of a liver phantom and simulated some intraoperative DTS projection images using a spherical ellipse scan geometry. Our results show a great improvement in the DTS image quality with the proposed method and prove the importance of choosing a good starting point for the iterative DTS reconstruction.

Digital tomosynthesis (DTS) is a 3-D imaging technique with limited depth resolution and can be used to guide a catheter during interventional lung procedures. DTS uses projection data acquired over a limited angular range to create tomographic slices with high in-plane resolution. This limited angular acquisition results in out-of-plane artifacts in the reconstructed volume. This work focused on improving the image quality of the tomosynthesis volume for lung imaging by selecting an acquisition trajectory that increases depth resolution, provides a 2-D blur response, and simplifies analytical reconstruction. We proposed a new pseudo-square acquisition geometry for DTS and performed the quantitative and qualitative comparison with arc (pseudo-linear) and circle acquisition geometries. We reconstructed DTS volumes for various angular ranges and numbers of projections for arc, circle, and pseudo-square trajectories using filtered back-projection and compared them with reference computed tomography (CT) scan using Pearson correlation coefficient (PCC). Based on simulations, we showed that pseudo-square acquisition resulted in better image quality than inbounded arc and inbounded circle geometries. It provided on average 5.5% and 26% higher PCC with reference CT than circle and arc DTS, respectively.

Data truncation is a common problem in computed tomography (CT). Truncation causes cupping artifacts inside the field-of-view (FOV) and anatomical structures missing outside the FOV. Deep learning has achieved impressive results in CT reconstruction from limited data. However, its robustness is still a concern for clinical applications. Although the image quality of learning-based compensation schemes may be inadequate for clinical diagnosis, they can provide prior information for more accurate extrapolation than conventional heuristic extrapolation methods. With extrapolated projection, a conventional image reconstruction algorithm can be applied to obtain a final reconstruction. In this work, a general plug-and-play (PnP) method for truncation correction is proposed based on this idea, where various deep learning methods and conventional reconstruction algorithms can be plugged in. Such a PnP method integrates data consistency for measured data and learned prior image information for truncated data. This shows to have better robustness and interpretability than deep learning only. To demonstrate the efficacy of the proposed PnP method, two state-of-the-art deep learning methods, FBPConvNet and Pix2pixGAN, are investigated for truncation correction in cone-beam CT in noise-free and noisy cases. Their robustness is evaluated by showing false negative and false positive lesion cases. With our proposed PnP method, false lesion structures are corrected for both deep learning methods. For FBPConvNet, the root-mean-square error (RMSE) inside the FOV can be improved from 92HU to around 30HU by PnP in the noisy case. Pix2pixGAN solely achieves better image quality than FBPConvNet solely for truncation correction in general. PnP further improves the RMSE inside the FOV from 42HU to around 27HU for Pix2pixGAN. The efficacy of PnP is also demonstrated on real clinical head data.

Three-dimensional cone-beam imaging has become valuable in interventional radiology. Currently, this tool, referred to as C-arm CT, employs a circular short-scan for data acquisition, which limits the axial volume coverage and yields unavoidable cone-beam artifacts. To improve flexibility in axial coverage and image quality, there is a critical need for novel data acquisition geometries and related image reconstruction algorithms. For this purpose, we previously introduced the extended line-ellipse-line trajectory, which allows complete scanning of arbitrary volume lengths in the axial direction together with adjustable axial beam collimation, from narrow to wide depending on the targeted application. A first implementation of this trajectory on a state-of-the-art robotic angiography system is reported here. More specifically, an assessment of the quality of this first implementation is presented. The assessment is in terms of geometric fidelity and repeatability, complemented with a first visual inspection of how well the implementation enables imaging an anthropomorphic head phantom. The geometric fidelity analysis shows that the ideal trajectory is closely emulated, with only minor deviations that have no impact on data completeness and clinical practicality. Also, mean backprojection errors over short-term repetitions are shown to be below the detector pixel size at field-of-view center for most views, which indicates repeatability is satisfactory for clinical utilization. These repeatability observations are further supported by values of the Structural Similarity Index Metric above 94% for reconstructions of the FORBILD head phantom from computer-simulated data based on repeated data acquisition geometries. Last, the real data experiment with the anthropomorphic head phantom shows that the high contrast features of the phantom are well reconstructed without distortions as well as without breaks or other disturbing transition zones, which was not obvious given the complexity of the data acquisition geometry and the major variations in axial coverage that occur over the scan.

Image reconstruction from insufficient data is common in computed tomography (CT), e.g., image reconstruction from truncated data, limited-angle data and sparse-view data. Deep learning has achieved impressive results in this field. However, the robustness of deep learning methods is still a concern for clinical applications due to the following two challenges: a) With limited access to sufficient training data, a learned deep learning model may not generalize well to unseen data; b) Deep learning models are sensitive to noise. Therefore, the quality of images processed by neural networks only may be inadequate. In this work, we investigate the robustness of deep learning in CT image reconstruction by showing false negative and false positive lesion cases. Since learning-based images with incorrect structures are likely not consistent with measured projection data, we propose a data consistent reconstruction (DCR) method to improve their image quality, which combines the advantages of compressed sensing and deep learning: First, a prior image is generated by deep learning. Afterwards, unmeasured projection data are inpainted by forward projection of the prior image. Finally, iterative reconstruction with reweighted total variation regularization is applied, integrating data consistency for measured data and learned prior information for missing data. The efficacy of the proposed method is demonstrated in cone-beam CT with truncated data, limited-angle data and sparse-view data, respectively. For example, for truncated data, DCR achieves a mean root-mean-square error of 24 HU and a mean structure similarity index of 0.999 inside the field-of-view for different patients in the noisy case, while the state-of-the-art U-Net method achieves 55 HU and 0.995 respectively for these two metrics.

We present further progress on the implementation of C-arm CT imaging with the extended line-ellipse-line (LEL) trajectory. This novel data acquisition geometry is designed to enhance image quality in interventional radiology. Previously, we showed that robust extended LEL data acquisition is feasible using a state-of-the-art multi-axis robotic C-arm (ARTIS pheno, Siemens Healthcare GmbH, Germany) and we also showed that accurate reconstruction from real data can be obtained using an iterative algorithm. The extensive computational effort required by such an algorithm is however not suitable for clinical translation. Reconstruction using a filtered-backprojection (FBP) formula would be practical. To use such a formula, there needs to be a technique to handle imperfections in the data acquisition geometry, which result from mechanical vibrations and gravity effects. We recently presented such a technique, but this development was only carried out for a single cycle of the LEL trajectory. In this work, we address the more challenging issue of reconstructing the volume covered by multiple cycles of the trajectory. Specifically, we propose an extension of our single cycle approach to multiple cycles. We successfully demonstrate that our procedure now allows seamless volume reconstruction from real data using a cone-beam performance phantom as well as an anthropomorphic head phantom. Our results bring the extended LEL trajectory closer to clinical deployment for improved image quality in interventional radiology. Further work will focus on increasing the number of views to avoid few view artifacts and on thoroughly demonstrating image quality benefits.

C-arm CT imaging can be improved in terms of axial coverage and cone-beam artifacts using advanced data acquisition geometries such as the extend line-ellipse-line trajectory. Previously, we showed that such a geometry can be robustly implemented on a clinical system. Here, we demonstrate that imperfections in the trajectory realization can be addressed so as to achieve accurate high contrast imaging with a theoretical-exact filtered- backprojection algorithm. The performance of the proposed algorithm is evaluated using the FORBILD head phantom as well as real data of an anthropomorphic head phantom.

Robustness of deep learning methods for limited angle tomography is challenged by two major factors: a) due to insufficient training data the network may not generalize well to unseen data; b) deep learning methods are sensitive to noise. Thus, generating reconstructed images directly from a neural network appears inadequate. We propose to constrain the reconstructed images to be consistent with the measured projection data, while the unmeasured information is complemented by learning based methods. For this purpose, a data consistent artifact reduction (DCAR) method is introduced: First, a prior image is generated from an initial limited angle reconstruction via deep learning as a substitute for missing information. Afterwards, a conventional iterative reconstruction algorithm is applied, integrating the data consistency in the measured angular range and the prior information in the missing angular range. This ensures data integrity in the measured area, while inaccuracies incorporated by the deep learning prior lie only in areas where no information is acquired. The proposed DCAR method achieves significant image quality improvement: for 120-degree cone-beam limited angle tomography more than 10% RMSE reduction in noise-free case and more than 24% RMSE reduction in noisy case compared with a state-of-the-art U-Net based method.

The application of traditional machine learning techniques, in the form of regression models based on conventional, “hand-crafted” features, to artifact reduction in limited angle tomography is investigated.

In computed tomography, image reconstruction from an insufficient angular range of projection data is called limited angle tomography. Due to missing data, reconstructed images suffer from artifacts, which cause boundary distortion, edge blurring, and intensity biases. Recently, deep learning methods have been applied very successfully to this problem in simulation studies. However, the robustness of neural networks for clinical applications is still a concern. It is reported that most neural networks are vulnerable to adversarial examples. In this paper, we aim to investigate whether some perturbations or noise will mislead a neural network to fail to detect an existing lesion. Our experiments demonstrate that the trained neural network, specifically the U-Net, is sensitive to Poisson noise. While the observed images appear artifact-free, anatomical structures may be located at wrong positions, e.g. the skin shifted by up to 1 cm. This kind of behavior can be reduced by retraining on data with simulated Poisson noise. However, we demonstrate that the retrained U-Net model is still susceptible to adversarial examples. We conclude the paper with suggestions towards robust deep-learning-based reconstruction.

In this work, the application of traditional machine learning techniques, in the form of regression models based on conventional, “hand-crafted” features, to streak reduction in limited angle tomography is investigated. Specifically, linear regression (LR), multi-layer perceptron (MLP), and reduced-error pruning tree (REPTree) are investigated. When choosing the mean-variation-median (MVM), Laplacian, and Hessian features, REPTree learns streak artifacts best and reaches the smallest root-mean-square error (RMSE) of 29HU for the Shepp-Logan phantom. Further experiments demonstrate that the MVM and Hessian features complement each other, whereas the Laplacian feature is redundant in the presence of MVM. Preliminary experiments on clinical data suggests that further investigation of clinical applications using REPTree may be worthwhile.

In previous work, we proposed a novel data acquisition geometry, called the Extended LEL trajectory, for Carm CT imaging in interventional radiology. This novel geometry aims at enabling larger axial field-of-view coverage without conebeam artifacts for imaging with a full X-ray beam as well as with a collimated X-ray beam used for scatter reduction purposes. In this work, we report on a first implementation of the Extended LEL trajectory on a state-of-the-art C-arm system. Highly satisfactory results are shown in terms of trajectory fidelity and repeatability. Suitability of the data for head imaging is also demonstrated using a Rando head phantom without and with 50% beam collimation.

The Papoulis-Gerchberg (P-G) algorithm is widely used for extrapolation of band-limited signals. It is applicable to limited angle tomography as well since typical imaged objects in computed tomography have a limited spatial extent, which means that the Fourier transforms of the objects can be considered band-limited signals. In computed tomography, some other bandlimitation properties have been discovered as well, which are referred to as data consistency conditions. For example, the Fourier transform of a parallel-beam sinogram has an empty double-wedge region. The Chebyshev-Fourier transform of a parallel-beam sinogram only has nonzero values inside a wedge region and these values form a checkerboard pattern, which is Helgason-Ludwig consistency condition. In this paper, we propose two P-G algorithms to restore missing data in limited angle tomography using the above two consistency conditions. Numerical experiments on the Shepp-Logan phantom demonstrate that they can reduce streaks better than the conventional P-G algorithm.

This paper addresses streak reduction in limited angle tomography. Although the iterative reweighted total variation (wTV) algorithm reduces small streaks well, it is rather inept at eliminating large ones since total variation (TV) regularization is scale-dependent and may regard these streaks as homogeneous areas. Hence, the main purpose of this paper is to reduce streak artifacts at various scales. We propose the scale-space anisotropic total variation (ssaTV) algorithm in two different implementations. The first implementation (ssaTV-1) utilizes an anisotropic gradient-like operator which uses 2s neighboring pixels along the streaks' normal direction at each scale s. The second implementation (ssaTV-2) makes use of anisotropic down-sampling and up-sampling operations, similarly oriented along the streaks' normal direction, to apply TV regularization at various scales. Experiments on numerical and clinical data demonstrate that both ssaTV algorithms reduce streak artifacts more effectively and efficiently than wTV, particularly when using multiple scales.

Tomographic reconstruction of cardiovascular structures from rotational angiograms acquired with interventional C-arm devices is challenging due to cardiac motion. Gating strategies are widely used to reduce data inconsistency but come at the cost of angular undersampling. We employ a spatio-temporally regularized 4-D reconstruction model, which is solved using a proximal algorithm, to handle the substantial undersampling associated with a strict gating setup. In a numerical phantom study based on the CAVAREV framework, similarity to the ground truth is improved from 82.3% to 87.6%by this approach compared to a state-of-the-art motion compensation algorithm, whereas previous regularized methods evaluated on this phantom achieved results below 80%. We also show first image results for a clinical patient data set.

Cone-beam (CB) computed tomography (CT) using a flooror ceiling-mounted C-arm system or using a robotic C-arm system has become a valuable tool in interventional radiology. This technology is typically used with a circular shortscan (SS) data acquisition geometry. As it is well-known, this geometry does not provide complete data for exact reconstruction. Furthermore, the classical SS-FDK algorithm, which is often employed, exacerbates CB artifacts by applying data redundancy weights that are not exact. In this paper, we are interested in mitigating CB artifacts in C-arm CT imaging of the head. We suggest a rebinning algorithm that allows transforming data from such systems into data acquired in an ideal geometry. This rebinning algorithm enables utilization of advanced CB reconstruction methods to mitigate CB artifacts. We extensively evaluated the performance of the rebinning algorithm using the FORBILD head phantom as well as a cylindrical phantom to assess MTF and SSP. The results show strong performance of the rebinning algorithm for geometry deviations encountered in practice. Furthermore, we have applied the rebinning algorithm to real data of an anthropomorphic phantom. This additional experiment demonstrated clinical value of our rebinning algorithm, particularly as a strong mitigation of CB artifacts can be seen in comparison with the SS-FDK algorithm. Last, we showed that data acquisition with the equivalent of a gantry tilt may provide, in terms of CB artifacts, additional improvements within the brain region.

In limited angle tomography, missing data in an insufficient angular scan will cause streak artifacts in the reconstructed images. Correspondingly, in the frequency domain representation of the imaged object, a double wedge-shaped region is missing. In this paper, we perform a regression in sinogram domain and an image fusion in frequency domain to restore the missing data. We first convert the sinogram restoration problem into a regression problem based on the Helgason–Ludwig consistency conditions. Due to its severe ill-posedness, regression only partially recovers the correct frequency components, especially lower frequency components, and will introduce erroneous ones, particularly higher frequencies. Bilateral filtering is utilized to retain the most prominent high frequency components and suppress erroneous ones. Afterwards, a fusion in the frequency domain utilizes the restored frequency components to fill the missing double wedge region. The proposed method is evaluated in a parallel-beam study on both numerical and clinical phantoms. The root-mean-square errors of the reconstructed images decrease from 302 to 78 HU for the noise-free Shepp–Logan phantom, from 355 to 175 HU for the noisy Shepp–Logan phantom, and from 187 to 56 HU for the clinical data. The results show that our method is promising in streak reduction and intensity offset compensation in both noise-free and noisy situations.

In limited angle tomography, only a limited angular range of data is acquired and consequently a double wedgeshaped region in the frequency domain representation of the imaged object is missing. Hence, streak artifacts occur. To restore the missing data, we perform a regression and an image fusion in sinogram domain and frequency domain, respectively. We first convert the sinogram restoration problem into a regression problem based on the Helgason-Ludwig consistency conditions. Due to the severe ill-posedness of the problem, regression only partially recovers the correct frequency components, especially lower frequency components, and will introduce erroneous ones, particularly higher frequencies. Bilateral filtering is utilized to retain the most prominent high frequency components and suppress erroneous ones. A fusion of the filtered image and the image reconstructed from the limited angle sinogram is performed afterwards in the frequency domain. The proposed method is evaluated on the Shepp-Logan phantom, for which the root-mean-square error of the reconstructed image decreases from 310 HU to 136 HU.

Dynamic cone-beam computed tomography (CBCT) imaging of the thorax, i. e. time-resolved reconstruction w. r. t. cardiac or respiratory motion, requires sophisticated algorithms, many of which are iterative and computationally expensive in terms of both runtime and memory. For the latter, hardware constraints pose a considerable challenge insofar as the volume grid cannot be chosen arbitrarily large. On the other hand, choosing a small grid may lead to severe artifacts if the object exceeds the size of the reconstruction domain. Additionally, lateral truncation of the projection data is commonly encountered as, e. g., flat panel detectors employed in interventional C-arm devices are not large enough to simultaneously image the entire width of the thorax in most patients. In iterative reconstruction, mild data truncation artifacts can also be alleviated by reconstructing on a sufficiently large grid. We present a simple model to incorporate information from outside the target grid in dynamic reconstruction. Its main component is the reconstruction of a static background image used to precompute an additive data correction term, which can be used in combination with any dynamic iterative reconstruction method. The effectiveness of our approach is demonstrated in a numerical phantom and clinical patient data.