Improved tomographic reconstruction of large-scale real-world data by filter optimization

Abstract

In advanced tomographic experiments, large detector sizes and large numbers of acquired datasets can make it difficult to process the data in a reasonable time. At the same time, the acquired projections are often limited in some way, for example having a low number of projections or a low signal-to-noise ratio. Direct analytical reconstruction methods are able to produce reconstructions in very little time, even for large-scale data, but the quality of these reconstructions can be insufficient for further analysis in cases with limited data. Iterative reconstruction methods typically produce more accurate reconstructions, but take significantly more time to compute, which limits their usefulness in practice. In this paper, we present the application of the SIRT-FBP method to large-scale real-world tomographic data. The SIRT-FBP method is able to accurately approximate the simultaneous iterative reconstruction technique (SIRT) method by the computationally efficient filtered backprojection (FBP) method, using precomputed experiment-specific filters. We specifically focus on the many implementation details that are important for application on large-scale real-world data, and give solutions to common problems that occur with experimental data. We show that SIRT-FBP filters can be computed in reasonable time, even for large problem sizes, and that precomputed filters can be reused for future experiments. Reconstruction results are given for three different experiments, and are compared with results of popular existing methods. The results show that the SIRT-FBP method is able to accurately approximate iterative reconstructions of experimental data. Furthermore, they show that, in practice, the SIRT-FBP method can produce more accurate reconstructions than standard direct analytical reconstructions with popular filters, without increasing the required computation time.

Unveiling water dynamics in fuel cells from time-resolved tomographic microscopy data

“…To expand the possible applications for large-scale datasets, improvements in computational efficiency are foreseen. For accelerated reconstruction, an interesting possibility is to exploit approximated algebraic filters [40][41][42] to obtain reconstructions with comparable image quality to SIRT with the computational effort of FBP. Moreover, further improvement of the efficiency of the timeregularization step is envisaged through neural networks which have already demonstrated impressive results in improving the tomographic reconstruction quality when used as a post-processing tool 43,44 .…”

Deep learning based classification of dynamic processes in time-resolved X-ray tomographic microscopy

“…A common analytical reconstruction algorithm for solving the tomographic inverse problem is FBP 35 . In FBP, the measured projection data is convolved with a filter prior to the back-projection step, defined 34 , 36 by where denotes a convolution of the filter with the measured data , and is the back-projection operator. The filter can be chosen depending on the SNR of the measured data to optimize the spatial resolution and contrast of the reconstruction.…”

“…The Simultaneous Iterative Reconstruction Technique (SIRT) 37 , 38 is an algebraic iterative reconstruction algorithm that models the tomographic linear system of equations. Starting from an initial reconstruction , typically a zero vector, the SIRT algorithm updates the reconstruction at each iteration by where is a relaxation factor influencing the convergence rate 36 . The algorithm is known to converge to a solution of …”

“…Once the filter is calculated, the same filter can be though applied to any other dataset that has the same number of projections, angular range and desired number of SIRT iterations in a computation time of a standard FBP reconstruction. In 2017, Pelt and Andrade 36 further demonstrated successful application of the algebraic filter approximation approach to several real-world large tomographic datasets, combining the filter both with FBP and Gridrec algorithms. A more detailed mathematical derivation of the filter and its performance characterization are presented in 34 , 36 .…”

Probing Novel Microstructural Evolution Mechanisms in Aluminum Alloys Using 4D Nanoscale Characterization

“…Python based toolbox used to analyze synchrotron tomography data [36,37] , was used to perform 3D reconstructions [38,39] . The stack of tomograms was post-processed using a combination of Mean 3D, Bandpass and Non-local means denoise filters in ImageJ [40] .…”