PhD defence Leise Borg

On 28 November, Leise Borg will defend her PhD thesis.

Title

Analysis of synchrotron X-ray tomographic data - Reconstruction and application.

Abstract

Synchrotron X-ray computed tomography (SXCT) allows to non-destructively image the interior of  objects at a very high resolution. Users of synchrotron facilities depend on reliable reconstructions (images) of the specimen they brought and subsequent analysis of the images. My dissertation addresses both of these issues.

Many in situ SXCT experiments entail rigs to control the in situ environment. This often leads to missing data, causing artifacts in the reconstructions, which complicates the subsequent image analysis. We analysed a specific tomographic experiment including a percolation cell for controlling fluid flow in chalk samples. The reconstructions suffered from severe artifacts. Hence, we developed methods to suppress the artifacts, and we generalized the methods by mathematically modeling the geometrical setup. The work was followed by a theoretical characterization of reconstruction artifacts from arbitrary incomplete SXCT data.

Developing image analysis in SXCT images was the focus of the other project. Highly specialized methods are often required to fully exploit the potentials of the nature of SXCT images. We analysed the muscle fiber morphology in 3D images of muscle biopsies taken from healthy participants and participants with cerebral palsy. Generally, morphology analyses are carried out in 2D images, but we showed that 3D image analysis provides a more accurate morphological description. Further, we introduced a measure of the orientation consistency of the muscle fibers, which is a measure of how aligned the fibers are in a sample. Interestingly, the orientation consistency was significantly larger for the healthy participants than for the participants with cerebral palsy.

Assessment Committee
Chairman: Associate Professor Kim Steenstrup Pedersen, Department of Computer Science, University of Copenhagen, Denmark
Reader in Applied and Computational Analysis Carola-Bibiane Schönlieb, University of Cambridge, UK
Associate Professor Martin Skovgaard Andersen, Technical University of Denmark, Denmark

Academic supervisor:
Professor Jon Sporring, Department of Computer Science, University of Copenhagen, Denmark

For an electronic copy of the thesis, please contact phdadmin@di.ku.dk