DART: a practical reconstruction algorithm for discrete tomography

A distributed ASTRA toolbox

While iterative reconstruction algorithms for tomography have several advantages compared to standard backprojection methods, the adoption of such algorithms in large-scale imaging facilities is still limited, one of the key obstacles being their high computational load. Although GPU-enabled computing clusters are, in principle, powerful enough to carry out iterative reconstructions on large datasets in reasonable time, creating efficient distributed algorithms has so far remained a complex task, requiring low-level programming to deal with memory management and network communication. The ASTRA toolbox is a software toolbox that enables rapid development of GPU accelerated tomography algorithms. It contains GPU implementations of forward and backprojection operations for many scanning geometries, as well as a set of algorithms for iterative reconstruction. These algorithms are currently limited to using GPUs in a single workstation. In this paper, we present an extension of the ASTRA toolbox and its Python interface with implementations of forward projection, backprojection and the SIRT algorithm that can be distributed over multiple GPUs and multiple workstations, as well as the tools to write distributed versions of custom reconstruction algorithms, to make processing larger datasets with ASTRA feasible. As a result, algorithms that are implemented in a high-level conceptual script can run seamlessly on GPU-enabled computing clusters, up to 32 GPUs or more. Our approach is not limited to slice-based reconstruction, facilitating a direct portability of algorithms coded for parallel-beam synchrotron tomography to cone-beam laboratory tomography setups without making changes to the reconstruction algorithm.

Fast and flexible X-ray tomography using the ASTRA toolbox

Object reconstruction from a series of projection images, such as in computed tomography (CT), is a popular tool in many different application fields. Existing commercial software typically provides sufficiently accurate and convenient-to-use reconstruction tools to the end-user. However, in applications where a non-standard acquisition protocol is used, or where advanced reconstruction methods are required, the standard software tools often are incapable of computing accurate reconstruction images. This article introduces the ASTRA Toolbox. Aimed at researchers across multiple tomographic application fields, the ASTRA Toolbox provides a highly efficient and highly flexible open source set of tools for tomographic projection and reconstruction. The main features of the ASTRA Toolbox are discussed and several use cases are presented.

Biological application of Compressed Sensing Tomography in the Scanning Electron Microscope

The three-dimensional tomographic reconstruction of a biological sample, namely collagen fibrils in human dermal tissue, was obtained from a set of projection-images acquired in the Scanning Electron Microscope. A tailored strategy for the transmission imaging mode was implemented in the microscope and proved effective in acquiring the projections needed for the tomographic reconstruction. Suitable projection alignment and Compressed Sensing formulation were used to overcome the limitations arising from the experimental acquisition strategy and to improve the reconstruction of the sample. The undetermined problem of structure reconstruction from a set of projections, limited in number and angular range, was indeed supported by exploiting the sparsity of the object projected in the electron microscopy images. In particular, the proposed system was able to preserve the reconstruction accuracy even in presence of a significant reduction of experimental projections.

Reduced projection angles for binary tomography with particle aggregation

This paper extends particle aggregate reconstruction technique (PART), a reconstruction algorithm for binary tomography based on the movement of particles. PART supposes that pixel values are particles, and that particles diffuse through the image, staying together in regions of uniform pixel value known as aggregates. In this work, a variation of this algorithm is proposed and a focus is placed on reducing the number of projections and whether this impacts the reconstruction of images. The algorithm is tested on three phantoms of varying sizes and numbers of forward projections and compared to filtered back projection, a random search algorithm and to SART, a standard algebraic reconstruction method. It is shown that the proposed algorithm outperforms the aforementioned algorithms on small numbers of projections. This potentially makes the algorithm attractive in scenarios where collecting less projection data are inevitable.

ICON: 3D reconstruction with 'missing-information' restoration in biological electron tomography

Electron tomography (ET) plays an important role in revealing biological structures, ranging from macromolecular to subcellular scale. Due to limited tilt angles, ET reconstruction always suffers from the 'missing wedge' artifacts, thus severely weakens the further biological interpretation. In this work, we developed an algorithm called Iterative Compressed-sensing Optimized Non-uniform fast Fourier transform reconstruction (ICON) based on the theory of compressed-sensing and the assumption of sparsity of biological specimens. ICON can significantly restore the missing information in comparison with other reconstruction algorithms. More importantly, we used the leave-one-out method to verify the validity of restored information for both simulated and experimental data. The significant improvement in sub-tomogram averaging by ICON indicates its great potential in the future application of high-resolution structural determination of macromolecules in situ.

A modified discrete algebraic reconstruction technique for multiple grey image reconstruction for limited angle range tomography

The `missing wedge', which is due to a restricted rotation range, is a major challenge for quantitative analysis of an object using tomography. With prior knowledge of the grey levels, the discrete algebraic reconstruction technique (DART) is able to reconstruct objects accurately with projections in a limited angle range. However, the quality of the reconstructions declines as the number of grey levels increases. In this paper, a modified DART (MDART) was proposed, in which each independent region of homogeneous material was chosen as a research object, instead of the grey values. The grey values of each discrete region were estimated according to the solution of the linear projection equations. The iterative process of boundary pixels updating and correcting the grey values of each region was executed alternately. Simulation experiments of binary phantoms as well as multiple grey phantoms show that MDART is capable of achieving high-quality reconstructions with projections in a limited angle range. The interesting advancement of MDART is that neither prior knowledge of the grey values nor the number of grey levels is necessary.

Quantitative 3D analysis of huge nanoparticle assemblies

Nanoparticle assemblies can be investigated in 3 dimensions using electron tomography. However, it is not straightforward to obtain quantitative information such as the number of particles or their relative position. This becomes particularly difficult when the number of particles increases. We propose a novel approach in which prior information on the shape of the individual particles is exploited. It improves the quality of the reconstruction of these complex assemblies significantly. Moreover, this quantitative Sparse Sphere Reconstruction approach yields directly the number of particles and their position as an output of the reconstruction technique, enabling a detailed 3D analysis of assemblies with as many as 10,000 particles. The approach can also be used to reconstruct objects based on a very limited number of projections, which opens up possibilities to investigate beam sensitive assemblies where previous reconstructions with the available electron tomography techniques failed.

TVR-DART: A More Robust Algorithm for Discrete Tomography From Limited Projection Data With Automated Gray Value Estimation

In this paper, we present a novel iterative reconstruction algorithm for discrete tomography (DT) named total variation regularized discrete algebraic reconstruction technique (TVR-DART) with automated gray value estimation. This algorithm is more robust and automated than the original DART algorithm, and is aimed at imaging of objects consisting of only a few different material compositions, each corresponding to a different gray value in the reconstruction. By exploiting two types of prior knowledge of the scanned object simultaneously, TVR-DART solves the discrete reconstruction problem within an optimization framework inspired by compressive sensing to steer the current reconstruction toward a solution with the specified number of discrete gray values. The gray values and the thresholds are estimated as the reconstruction improves through iterations. Extensive experiments from simulated data, experimental μCT, and electron tomography data sets show that TVR-DART is capable of providing more accurate reconstruction than existing algorithms under noisy conditions from a small number of projection images and/or from a small angular range. Furthermore, the new algorithm requires less effort on parameter tuning compared with the original DART algorithm. With TVR-DART, we aim to provide the tomography society with an easy-to-use and robust algorithm for DT.

Binary tomography reconstruction from few projections with Total Variation regularization for bone microstructure studies

Discrete tomography refers to a class of reconstruction methods adapted to discrete-valued images. Many different approaches have been investigated to address the binary case, when a two-phase object is considered. This reconstruction problem is very important in medical or material applications where it is crucial to reduce the number of projections. In this paper, we address the problem of binary image reconstruction for X-ray CT imaging from a small number of projections. We propose a TV (Total Variation) regularization approach and compare the results obtained with or without an additional box convex constraint. The schemes are applied to a simple disk image and to more complex bone cross-sections for various noise levels. The minimization of the regularization functional is performed with the state-of-the-art ADMM (Alternate Direction Minimization Method) algorithm. The methods perform equally well on a simple disk image. The additional box convex constraints improves the reconstruction results for complex structures with fine details.

Rapid low dose electron tomography using a direct electron detection camera

We demonstrate the ability to record a tomographic tilt series containing 3487 images in only 3.5 s by using a direct electron detector in a transmission electron microscope. The electron dose is lower by at least one order of magnitude when compared with that used to record a conventional tilt series of fewer than 100 images in 15-60 minutes and the overall signal-to-noise ratio is greater than 4. Our results, which are illustrated for an inorganic nanotube, are important for ultra-low-dose electron tomography of electron-beam-sensitive specimens and real-time dynamic electron tomography of nanoscale objects with sub-ms temporal resolution.

A matter of quantum voltages

Voltages inside matter are relevant to crystallization, materials science, biology, catalysis, and aqueous chemistry. The variation of voltages in matter can be measured by experiment, however, modern supercomputers allow the calculation of accurate quantum voltages with spatial resolutions of bulk systems well beyond what can currently be measured provided a sufficient level of theory is employed. Of particular interest is the Mean Inner Potential (V(o))--the spatial average of these quantum voltages referenced to the vacuum. Here we establish a protocol to reliably evaluate V(o) from quantum calculations. Voltages are very sensitive to the distribution of electrons and provide metrics to understand interactions in condensed phases. In the present study, we find excellent agreement with measurements of V(o) for vitrified water and salt crystals and demonstrate the impact of covalent and ionic bonding as well as intermolecular/atomic interactions. Certain aspects in this regard are highlighted making use of simple model systems/approximations. Furthermore, we predict V(o) as well as the fluctuations of these voltages in aqueous NaCl electrolytes and characterize the changes in their behavior as the resolution increases below the size of atoms.

Graph-cut based discrete-valued image reconstruction

Efficient graph-cut methods have been used with great success for labeling and denoising problems occurring in computer vision. Unfortunately, the presence of linear image mappings has prevented the use of these techniques in most discrete-amplitude image reconstruction problems. In this paper, we develop a graph-cut based framework for the direct solution of discrete amplitude linear image reconstruction problems cast as regularized energy function minimizations. We first analyze the structure of discrete linear inverse problem cost functions to show that the obstacle to the application of graph-cut methods to their solution is the variable mixing caused by the presence of the linear sensing operator. We then propose to use a surrogate energy functional that overcomes the challenges imposed by the sensing operator yet can be utilized efficiently in existing graph-cut frameworks. We use this surrogate energy functional to devise a monotonic iterative algorithm for the solution of discrete valued inverse problems. We first provide experiments using local convolutional operators and show the robustness of the proposed technique to noise and stability to changes in regularization parameter. Then we focus on nonlocal, tomographic examples where we consider limited-angle data problems. We compare our technique with state-of-the-art discrete and continuous image reconstruction techniques. Experiments show that the proposed method outperforms state-of-the-art techniques in challenging scenarios involving discrete valued unknowns.

An accurate projection model for diffraction image formation and inversion using a polychromatic cone beam

Recently, the concept of X-ray diffraction contrast tomography (DCT) has been extended to the case of more widely available laboratory source CT systems. Using well known concepts from geometrical ray optics, an exact formulation is derived for the forward and backward projection geometry encountered under polychromatic cone beam illumination, and it is shown how this projection model can be efficiently implemented in practice. The new projection model is subsequently used for iterative tomographic reconstruction of the three-dimensional shape of a grain from a set of experimentally observed cone beam projections and shows a clear improvement compared to the simplified projection model used previously.

A compressed sensing based approach on Discrete Algebraic Reconstruction Technique

Discrete tomography (DT) techniques are capable of computing better results, even using less number of projections than the continuous tomography techniques. Discrete Algebraic Reconstruction Technique (DART) is an iterative reconstruction method proposed to achieve this goal by exploiting a prior knowledge on the gray levels and assuming that the scanned object is composed from a few different densities. In this paper, DART method is combined with an initial total variation minimization (TvMin) phase to ensure a better initial guess and extended with a segmentation procedure in which the threshold values are estimated from a finite set of candidates to minimize both the projection error and the total variation (TV) simultaneously. The accuracy and the robustness of the algorithm is compared with the original DART by the simulation experiments which are done under (1) limited number of projections, (2) limited view problem and (3) noisy projections conditions.

A multiresolution approach to discrete tomography using DART

In discrete tomography, a scanned object is assumed to consist of only a few different materials. This prior knowledge can be effectively exploited by a specialized discrete reconstruction algorithm such as the Discrete Algebraic Reconstruction Technique (DART), which is capable of providing more accurate reconstructions from limited data compared to conventional reconstruction algorithms. However, like most iterative reconstruction algorithms, DART suffers from long computation times. To increase the computational efficiency as well as the reconstruction quality of DART, a multiresolution version of DART (MDART) is proposed, in which the reconstruction starts on a coarse grid with big pixel (voxel) size. The resulting reconstruction is then resampled on a finer grid and used as an initial point for a subsequent DART reconstruction. This process continues until the target pixel size is reached. Experiments show that MDART can provide a significant speed-up, reduce missing wedge artefacts and improve feature reconstruction in the object compared with DART within the same time, making its use with large datasets more feasible.

Autonomous reconstruction and segmentation of tomographic data

A Bayesian approach to reconstruction and segmentation of tomographic data is outlined and further detailed for the case of absorption tomography. The algorithm allows the quantification of reconstruction errors and segmentation confidence. Calculation results for various experimental settings (number of projections, incident dose, different materials) are shown and discussed.

Super-resolution for computed tomography based on discrete tomography

In computed tomography (CT), partial volume effects impede accurate segmentation of structures that are small with respect to the pixel size. In this paper, it is shown that for objects consisting of a small number of homogeneous materials, the reconstruction resolution can be substantially increased without altering the acquisition process. A super-resolution reconstruction approach is introduced that is based on discrete tomography, in which prior knowledge about the materials in the object is assumed. Discrete tomography has already been used to create reconstructions from a low number of projection angles, but in this paper, it is demonstrated that it can also be applied to increase the reconstruction resolution. Experiments on simulated and real μCT data of bone and foam structures show that the proposed method indeed leads to significantly improved structure segmentation and quantification compared with what can be achieved from conventional reconstructions.

Three-dimensional chemical mapping by EFTEM-TomoJ including improvement of SNR by PCA and ART reconstruction of volume by noise suppression

Electron tomography is becoming one of the most used methods for structural analysis at nanometric scale in biological and materials sciences. Combined with chemical mapping, it provides qualitative and semiquantitative information on the distribution of chemical elements on a given sample. Due to the current difficulties in obtaining three-dimensional (3D) maps by energy-filtered transmission electron microscopy (EFTEM), the use of 3D chemical mapping has not been widely adopted by the electron microscopy community. The lack of specialized software further complicates the issue, especially in the case of data with a low signal-to-noise ratio (SNR). Moreover, data interpretation is rendered difficult by the absence of efficient segmentation tools. Thus, specialized software for the computation of 3D maps by EFTEM needs to include optimized methods for image series alignment, algorithms to improve SNR, different background subtraction models, and methods to facilitate map segmentation. Here we present a software package (EFTEM-TomoJ, which can be downloaded from http://u759.curie.fr/fr/download/softwares/EFTEM-TomoJ), specifically dedicated to computation of EFTEM 3D chemical maps including noise filtering by image reconstitution based on multivariate statistical analysis. We also present an algorithm named BgART (for background removing algebraic reconstruction technique) allowing the discrimination between background and signal and improving the reconstructed volume in an iterative way.

Fast tomographic reconstruction from limited data using artificial neural networks

Image reconstruction from a small number of projections is a challenging problem in tomography. Advanced algorithms that incorporate prior knowledge can sometimes produce accurate reconstructions, but they typically require long computation times. Furthermore, the required prior knowledge can be very specific, limiting the type of images that can be reconstructed. Here, we present a reconstruction method that automatically learns prior knowledge using an artificial neural network. We show that this method can be viewed as a combination of filtered backprojection steps, and, therefore, has a relatively low computational cost. Results for two different cases show that the new method is able to use the learned information to produce high quality reconstructions in a short time, even when presented with a small number of projections.

Validation of three-dimensional diffraction contrast tomography reconstructions by means of electron backscatter diffraction characterization

Microstructure reconstructions resulting from diffraction contrast tomography data of polycrystalline bulk strontium titanate were reinvestigated by means of electron backscatter diffraction (EBSD) characterization. Corresponding two-dimensional grain maps from the two characterization methods were aligned and compared, focusing on the spatial resolution at the internal interfaces. The compared grain boundary networks show a remarkably good agreement both morphologically and in crystallographic orientation. Deviations are critically assessed and discussed in the context of diffraction data reconstruction and EBSD data collection techniques.

Automatic parameter estimation for the discrete algebraic reconstruction technique (DART)

Computed tomography (CT) is a technique for noninvasive imaging of physical objects. In the discrete algebraic reconstruction technique (DART), prior knowledge about the material's densities is exploited to obtain high quality reconstructed images from a limited number of its projections. In practice, this prior knowledge is typically not readily available. Here, a fully automatic method, called projection distance minimization DART (PDM-DART), is proposed in which the optimal grey level parameters are adaptively estimated during the reconstruction process. To apply PDM-DART, only the number of different grey levels should be known in advance. Simulation as well as real μCT experiments show that PDM-DART is capable of computing reconstructed images of which the quality is similar to reconstructions computed by conventional DART based on exact prior knowledge, thereby eliminating the need for tedious and error-prone user interaction.