Variability of MRI-based renal blood flow measurements

Abstract

Renal blood flow (RBF) and perfusion are valuable parameters that reflect important features of the renal function. Magnetic resonance imaging (MRI) enables the determination of RBF and perfusion using phase contrast (PC), arterial spin labelling (ASL), and dynamic contrast enhanced (DCE). The clinical utility of these MRI techniques is compromised by large variability owing to physiological and technical confounders. The main aim of this thesis is to identify and evaluate the sources of variability in MRI-based measures of RBF and perfusion.

A literature review was conducted to identify and classify the physiological confounders of RBF and perfusion, which informed the formulation of evidence-based recommendations for future studies. An assessment of different technical confounders was presented in terms of data acquisition and post-processing of MRI-based RBF and perfusion. Substantial differences were observed in DCE perfusion parameters, owing to multiple post-processing choices. Comparing RBF between PC and DCE acquired from two sites showed large dependence on the site and scanner type attributed to larger artefacts and noise in the older scanner model. A detailed assessment of the bias and precision of PC, ASL, and DCE provides insights into the technical performance of these methods. Intra- subject comparison showed small bias but poor agreement between all methods. Repeatability in healthy volunteers showed smaller error in PC than ASL, which supports the potential of PC as a useful tool to measure RBF. Reduced RBF was observed in the healthy cohort compared with reference literature value. An investigation into the effect of fasting was conducted to rule out the impact of patient preparation as a potential confounder. Results indicate no systematic effect of fasting on average MRI measurements of RBF and perfusion in healthy volunteers. The work in this thesis provides evidence to inform guidelines for renal imaging in research and clinical settings and further contributes to current standardisation efforts for quantification of RBF and perfusion using MRI.