Quantitative MRI for measuring fibrosis in heart and liver

Fibrosis is a hallmark characteristic of numerous pathologies. This pathological accumulation of connective tissue has the potential to lead to organ dysfunction due to the loss of functioning cells and changes in tissue structure. This thesis presents the development and implementation of various quantitative magnetic resonance imaging (MRI) techniques to detect alterations in the biomechanical or physiological properties of tissue caused by fibrosis. The aim was to improve the non-invasive evaluation and staging of fibrosis in the heart and liver.
Starting with magnetic resonance elastography (MRE), the gravitational transducer was developed to improve MRE image quality and was shown to generate high-quality tissue stiffness maps. To facilitate the clinical integration of MRE, a transducer-free MRE method using intrinsic cardiac motion for shear wave generation was devised. In the heart, this led to promising results, with cardiac patients showing significantly higher shear wave speeds than in healthy volunteers. However, application in the liver showed this technique was not suitable for use in chronic liver disease patients. Moving on to multiparametric MRI, the combination of multiple quantitative MRI techniques made it possible to assess the full spectrum of disease in non-alcoholic fatty liver disease. Finally, a neural network to fit a tri-exponential model to intravoxel incoherent motion (IVIM) data was successfully implemented, resulting in high-quality parameter maps.