Tomography is a powerful technique to non-destructively determine the interior structure of an object.Usually, a series of projection images (e.g.\ X-ray images) is acquired from a range of... Show moreTomography is a powerful technique to non-destructively determine the interior structure of an object.Usually, a series of projection images (e.g.\ X-ray images) is acquired from a range of different positions.from these projection images, a reconstruction of the object's interior is computed. Many advanced applications require fast acquisition, effectively limiting the number of projection images and imposing a level of noise on these images. These limitations result in artifacts (deficiencies) in the reconstructed images. Recently, deep neural networks have emerged as a powerful technique to remove these limited-data artifacts from reconstructed images, often outperformingconventional state-of-the-art techniques. To perform this task, the networks are typically trained on a dataset of paired low-quality and high-quality images of similar objects. This is a major obstacle to their use in many practical applications. In this thesis, we explore techniques to employ deep learning in advanced experiments where measuring additional objects is not possible. Show less
With tomography it is possible to reconstruct the interior of an object without destroying. It is an important technique for many applications in, e.g., science, industry, and medicine. The runtime... Show moreWith tomography it is possible to reconstruct the interior of an object without destroying. It is an important technique for many applications in, e.g., science, industry, and medicine. The runtime of conventional reconstruction algorithms is typically much longer than the time it takes to perform the tomographic experiment, and this prohibits the real-time reconstruction and visualization of the imaged object. The research in this dissertation introduces various techniques such as new parallelization schemes, data partitioning methods, and a quasi-3D reconstruction framework, that significantly reduce the time it takes to run conventional tomographic reconstruction algorithms without affecting image quality. The resulting methods and software implementations put reconstruction times in the same ballpark as the time it takes to do a tomographic scan, so that we can speak of real-time tomographic reconstruction. Show less
The research in this thesis is focused on tomographic reconstruction based on two imaging modalities in electron microscopy. The first modality is high angle annular dark field scanning... Show moreThe research in this thesis is focused on tomographic reconstruction based on two imaging modalities in electron microscopy. The first modality is high angle annular dark field scanning transmission microscopy (HAADF-STEM), and the second modality is energy-dispersive X-ray spectroscopy (EDS). In this Ph.D. thesis, we propose several approaches to pave the way for HAADF- STEM + EDS tomography: (1) the HAADF-EDS bimodal tomographic reconstruction technique, which is based on jointly modeling the consistency of the two imaging modalities; (2) TNV-regularized joined reconstruction which allows to incorporate the prior knowledge that common edges exist in the reconstructions from HAADF and EDS data respectively; (3) a set of algorithmic recipes to tailor various reconstruction algorithms for given experimental conditions and sample properties; (4) an algorithm for automatically correcting the nonlinear damping effects in HAADF-STEM tomographic data. Show less
Computed Tomography (CT) is an imaging technique that is used to calculate the interior of an object using X-rays under multiple projection angles. A well-known application is medical imaging with... Show moreComputed Tomography (CT) is an imaging technique that is used to calculate the interior of an object using X-rays under multiple projection angles. A well-known application is medical imaging with a CT-scanner. The reconstruction methods can roughly be divided into two categories: analytical reconstruction methods and algebraic reconstruction methods (ARMs). An example of an algorithm from the first category is Filtered Backprojection (FBP). This method has a high computational efficiency and it performs well in cases with many equiangularly distributed projection angles and high signal-to-noise ratio. ARMs require in general more computation time. They are more robust with respect to noise and can handle few projection angles or a limited angular range better. In this dissertation, the new algorithm Algebraic filter – Filtered Backprojection (AF-FBP) is introduced, which uses an ARM to create filters that can be used in FBP. The reconstruction quality of AF-FBP approximates that of the corresponding (locally) linear ARM, while the reconstructions are obtained with the computational efficiency of FBP. In cases with a small number of different scanning geometries, using AF-FBP enables the reconstruction of images of relatively high quality for few projection angles, limited angular range, or low signal-to-noise ratio. Show less
We present a computational approach for fast approximation of nonlinear tomographic reconstruction methods by filtered backprojection (FBP) methods. Algebraic reconstruction algorithms are the... Show moreWe present a computational approach for fast approximation of nonlinear tomographic reconstruction methods by filtered backprojection (FBP) methods. Algebraic reconstruction algorithms are the methods of choice in a wide range of tomographic applications, yet they require significant computation time, restricting their usefulness. We build upon recent work on the approximation of linear algebraic reconstruction methods and extend the approach to the approximation of nonlinear reconstruction methods which are common in practice. We demonstrate that if a blueprint image is available that is sufficiently similar to the scanned object, our approach can compute reconstructions that approximate iterative nonlinear methods, yet have the same speed as FBP. Show less
Gold nanoparticles are studied extensively due to their unique optical and catalytical properties. Their exact shape determines the properties and thereby the possible applications. Electron... Show moreGold nanoparticles are studied extensively due to their unique optical and catalytical properties. Their exact shape determines the properties and thereby the possible applications. Electron tomography is therefore often used to examine the three-dimensional (3D) shape of nanoparticles. However, since the acquisition of the experimental tilt series and the 3D reconstructions are very time consuming, it is difficult to obtain statistical results concerning the 3D shape of nanoparticles. Here, we propose a new approach for electron tomography that is based on artificial neural networks. The use of a new reconstruction approach enables us to reduce the number of projection images with a factor of 5 or more. The decrease in acquisition time of the tilt series and use of an efficient reconstruction algorithm allows us to examine a large amount of nanoparticles in order to retrieve statistical results concerning the 3D shape. Show less
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... Show moreIn 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. Show less
