NMR techniques for the characterisation of tissue remodelling

Since its introduction into clinical routine, magnetic resonance imaging (MRI) has become an essential imaging technique due to a distinguished soft tissue contrast and the absence of ionizing radiation. The quantitative analysis of MRI data permits to study tissue composition and to investigate tissue changes over time. Thereby it can be based on endogenous MR parameters like relaxation times or proton density but also on quantities inherently influenced by these parameters, like image contrast information.
This thesis presents different approaches of quantitative analysis of magnetic resonance data to address currently unsolved problems in the field of medical and clinical forensic imaging. For this purpose biomarkers linked to tissue changes are measured or newly derived using magnetic resonance to characterize tissue changes.
A study on estimating the age of subcutaneous soft tissue haematomas was solely based on image contrast information, which was analysed to create a model and provide an objective age estimation.
For the observation of bone bridge formation, contrast enhanced MRI data of fractures in an animal model were investigated. This information was then put into a model to derive pharmacokinetic parameters which led to an early detection of bone bridge formation before it was visible in morphologic imaging data.
In a spectroscopy-based study the age-related conversion of red to yellow bone marrow is investigated. By applying a dedicated acquisition protocol, relaxation times of human lumbar vertebrae are derived and analysed for their reproducibility and for their correlation with age.
Ultimately, the contrast of epiphyseal gaps in MR images of the hand is investigated in a study on the potential acceleration of MRI for forensic age estimation. In this study image raw data is undersampled retrospectively. The reconstructed images are then analysed by radiologists and a fully automatic method to derive the limit of acceleration that can safely be applied to MR imaging for forensic age estimation without influencing the outcome.