Quantitative analysis and measurement of flow using magnitude and phase magnetic resonance imaging.

Abstract

Magnetic Resonance Imaging (MRI) has the ability to image flow qualitatively and quantitatively. However, in the presence of complex flow states, the contrast mechanism in flow encoded MRI is not yet fully understood, and the validity of quantitative velocity measurements using MR phase imaging techniques needs to be established. Quantitative MRI flow studies were performed on a non-axially symmetrically obstructed tube phantom using a flow encoded MRI phase subtraction technique which generated both magnitude and phase images. On the phase images, regions of complex flow features including flow stagnation, separation and laminar flow, could be clearly identified. It was demonstrated that the source contributing to signal contrast in the magnitude images was an admixture of Time Of Flight (TOF) enhancement due to in-flow and signal void due to phase dispersion in the presence of high shear. A new MRI technique for flow measurement was developed which employed both TOF and phase effects and generated magnitude and phase images of a displaced bolus of spins in the direction of flow, for internal calibration of the velocity measurements. Very good correlation was established among velocity values obtained using the two different mechanisms for the range of 5 cm/s to 29 cm/s. MRI velocity measurement results were further validated by comparison with numerical solutions for three dimensional velocity distributions based on basic fluid dynamic equations, obtained using a finite difference scheme. The velocity profiles obtained from the simulation and MRI agreed well for a range of flow rates. A slight discrepency was likely caused by phase errors due to non-compensated higher order motion of the spins, eddy current effects, and undersampling effects. The first two factors could be more significant at high flow rates. A flow artifact due to movement of spins in the frequency encoding direction in MR images was demonstrated both theoretically, by simulating a model spin density distribution using both continuous and discrete Fourier Transform techniques, and experimentally. In general, this type of artifact had an oscillatory appearance whose extent might be enhanced or diminished when convolved with undersampling effects. This thesis research has increased our understanding of the underlying physical mechanisms and established the validity of MRI for quantitative flow measurements. The MRI flow encoded phase imaging technique is capable of measuring a flow field with fine spatial resolution in a short amount of time, rendering it a unique and powerful tool for both qualitative and quantitative flow studies.