Nuclear magnetic resonance and microcirculation: The influence of pulsatile brain-tissue motion on measurements of intravoxel incoherent motion and assessment of haemodynamics using exo- and endogenous tracers

Abstract: In this project, the potential of magnetic resonance (MR) imaging and spectroscopy in studies of microcirculation and haemodynamics was evaluated. The spatial and temporal characteristics of human pulsatile brain-tissue movements in healthy individuals, relevant for the understanding of the cerebrospinal-fluid (CSF) circulation and the pathogenesis of hydrocephalus, were thoroughly investigated. Abnormal motion patterns in brain tissue and CSF were demonstrated in patients with intracranial tumours, and navigator echoes were shown to reduce motion artifacts in phase images. Intravoxel phase dispersion, caused by a spatial distribution of brain-tissue velocities, was shown to act as a source of uncertainty in gradient-sensitised MR imaging of intravoxel incoherent motion (IVIM). The diffusion coefficient was considerably overestimated if measured during inappropriate periods in the cardiac cycle, a finding which may also prove important in IVIM-encoded echo-planar imaging (EPI). Using IVIM-sensitive EPI, the perfusion fraction in patients with acute ischaemic stroke was shown to be considerably smaller in the infarct region than in normal tissue in the contralateral hemisphere. In a deuterium MR spectroscopy study, the D2O washout in animals, after intratissue injection, displayed a pronounced injection-volume dependence as well as a non-physiological component of signal decay. The reproducibility was reasonable in subcutaneous tissue, but rather poor in heterogeneous tumours. In a dynamic susceptibility-contrast MR imaging study, the first passage of an exogenous tracer (Gd-DTPA- BMA) was monitored in the normal human brain, and reasonable regional cerebral blood volume (rCBV) values were observed. The rCBV and the perfusion fraction in white matter showed a promising tendency to correlate. Finally, functional MR imaging (fMRI) studies of cortical activation were initiated in a typical clinical environment, and it was verified that cortical activation is observable both at 1.0 T and 1.5 T. Investigations of the field-strength dependence might, to some extent, provide information on the origin of the fMRI contrast in gradient-echo imaging.