Modeling and Measurement of Tissue Compartments with Fast Signal Decay in Whole-Body Magnetic Resonance Imaging

Abstract

The aim of this thesis is to characterize the signal behavior of tissue components with fast signal decay in whole-body magnetic resonance imaging. Therefore, the first chapter contains a summary of the physical basics of magnetic resonance. In the following second chapter, the modeling of a possible water signal in inflamed adipose tissue is performed. For this purpose, the characteristic geometry of the water inclusions located between adipocytes is exploited to draw conclusions about the relaxation behavior and, thus, the measurability of the water signal by means of simulations and phantom measurements. It can be concluded that short echo times are necessary for detection of these fast-relaxing components. This is provided, for example, by a newly introduced method in the third chapter, which allows for the calculation of spectra based on specific imaging data sets with ultra-short echo times. Based on this approach, a characterization of the collagen signal is performed. With the resulting signal model, collagen in aqueous solution can be detected starting at a mass fraction of 2-4%. These values are attributable to a range which would be relevant, for example, for the staging of fibrosis in liver tissue. The method of calculating spatially resolved spectra using ultra-short echo times is also applied in the fourth chapter. In this chapter, modeling of the water signal originating from compact bone is performed. Thereby, different compartments can be distinguished. Usually, bound water is separated from free water present in porous structures. This typically results due to the different relaxation times of the compartments. In this section, it is postulated that the free water can be further subdivided into signal components with different frequencies based on the geometry of the shaping structures. This hypothesis is tested using the newly proposed method.