Towards quantitative magnetic resonance assessment in parenchymal liver disease

Open Access
Authors
  • J.H. Runge
Supervisors
Cosupervisors
  • A.J. Nederveen
Award date 13-02-2015
ISBN
  • 9789491602306
Number of pages 149
Organisations
  • Faculty of Medicine (AMC-UvA)
Abstract
In this thesis several advanced magnetic resonance (MR) techniques for quantitative measurements in parenchymal liver disease are studied. In particular, certain important hallmarks of liver disease such as steatosis, fibrosis, iron overload and inflammation are studied. Steatosis or fatty liver disease can be ascertained invasively—with biopsy—but is preferable assessed non-invasively, for example with MR techniques. This thesis shows that both MR imaging and spectroscopy are preferable over qualitative and quantitative (invasive) histological assessment in terms of correlation with biochemically determined liver fat content. Moreover, MR spectroscopy derived fat signal fractions can be confidently converted to fat mass fractions, allowing easier interpretation and comparison with biochemical analyses. MR elastography can be used for MR-based fibrosis assessment and is shown to have very good interobserver agreement, supporting its use as non-invasive screening tool for liver fibrosis. For liver iron content assessment MRI offers several options. This work shows that in terms of success rates, the Gandon and R2* methods are preferable over the R2 method. Moreover, as the R2* has better interobserver agreement, requires fewer breath holds and can distinguish a wider range of liver iron levels, it is the method of choice. For inflammation assessment, MR has no single, clear option yet. 31P MR spectroscopy is one of the possible methods and allows probing of phosphorus-containing liver metabolites. The MR spectroscopic imaging technique "AMESING" allows calculation of local T2 times and therewith the acquisition of localised, high-SNR spectra. In the future, these may be used to assess tissue inflammation.
Document type PhD thesis
Note Research conducted at: Universiteit van Amsterdam
Language English
Downloads
Permalink to this page
cover
Back