Volumetric MRI measurements of velocity and flow - Accuracy, visualisation and technical improvements

Sammanfattning: In the last two decades, phase-contrast magnetic resonance imaging (PC-MRI) has evolved from two-dimensional velocity and flow measurements to volumetric, time-resolved depictions of velocity fields (4D PC). The acquisition of time-resolved velocity fields allows flow visualisations that might provide better understanding of the dynamics of the cardiac system. Because 4D PC is increasingly being used to quantify certain physiological parameters and derive others, the need for validation and sufficient accuracy is increasing. Additionally, because 4D PC is a time-consuming technique, strategies for reducing the acquisition time are being developed and are crucial for 4D PC to realise its clinical potential. The aims of the papers in this thesis were the following: 1) To validate a conventional 4D PC sequence with two commonly used acquisition acceleration strategies in a phantom setup, especially with respect to the visualisation accuracy (Paper I). Furthermore, two common background phase correction strategies were compared. The results of this study showed that background phase correction is important for accurate flow visualisations and quantitative flow measurement. 2) develop and evaluate a 4D PC sequence with a spiral readout scheme and short echo time (TE) for the depiction of high velocities in restricted geometries, and to investigate its properties compared to those of standard approaches (Paper II). The results of this study showed that short-TE sequences with spiral readouts accurately quantified the maximum velocities, and demonstrated the feasibility of using volumetric sequences for complete coverage of the stenotic region. 3) To develop and investigate a 4D PC sequence with variable velocity encoding (4D vPC) throughout the cardiac cycle in an effort to reduce the noise and improve visualisations of time-varying flow patterns (Paper III). The conclusion drawn from this study was that variable velocity encoding provided lower noise levels in the diastolic parts of the cardiac cycle and improved flow visualisations. 4) To evaluate the previously constructed 4D vPC sequence with respect to wall shear stress (WSS) measurements in the aorta (Paper IV). The measured WSS agreed with the corresponding measurements performed with conventional 4D PC in both the systole and diastole. Furthermore, a smoother WSS distribution could be observed in the 4D vPC datasets.