title: Shearletbasierte  Suszeptibilitätskartenrekonstruktion mit  automatischer Parameterwahl
creator: Stiegeler, Janis
subject: ddc-530
subject: 530 Physics
subject: ddc-600
subject: 600 Technology (Applied sciences)
description: Quantitative susceptibility mapping (QSM) attempts to calculate the suscepti-  bility distribution of tissue using magnetic resonance techniques. The purpose  of this work is to find an image quality measure, which effectively prevents  an over-smoothing of the susceptibility maps, when optimizing the parameters  with regard to this measure. Moreover, an algorithm to solve the inverse phase  to susceptibility problem is introduced, which uses verified and mathematically  founded methods to compute the susceptibility distribution accurately. To achie-  ve this the shearlet transform, the generalized total variation and the method of  iterative re-weighting are used. Furthermore, a step is made towards application  of QSM in clinical practise, namely the automatic selection of the regularization  parameters. For this the method of unbiased predictive risk estimator is used.  The values of the image quality measures of the calculated susceptibility maps  achieve throughout top rankings in comparison with the submitted algorithms  at the 1th and 2nd QSM reconstruction challenge. Furthermore, the calculated  susceptibility maps are free from strong artefacts and their visual perception is  in accordance with its good image quality values. Probably the anisotropic pro-  perties of the shearlets lead to such an accurate calculation of local susceptibility  sources.
date: 2022
type: Dissertation
type: info:eu-repo/semantics/doctoralThesis
type: NonPeerReviewed
format: application/pdf
identifier: https://archiv.ub.uni-heidelberg.de/volltextserverhttps://archiv.ub.uni-heidelberg.de/volltextserver/31819/1/Dissertation.pdf
identifier: DOI:10.11588/heidok.00031819
identifier: urn:nbn:de:bsz:16-heidok-318192
identifier:   Stiegeler, Janis  (2022) Shearletbasierte Suszeptibilitätskartenrekonstruktion mit automatischer Parameterwahl.  [Dissertation]     
relation: https://archiv.ub.uni-heidelberg.de/volltextserver/31819/
rights: info:eu-repo/semantics/openAccess
rights: http://archiv.ub.uni-heidelberg.de/volltextserver/help/license_urhg.html
language: ger