The objective of this project is to develop ultra-fast magnetic resonance imaging techniques and applying them to measuring flow and motion. This is accomplished by: (i) developing ultra-fast imaging for assessment of simple linear (i.e., one directional) flow; (ii) developing ultra-fast imaging for application and assessment of more complex and other rapidly-varying phenomena; (iii) applying ultra-fast imaging techniques developed for Specific Aims (i) and (ii) to the analysis of in-vivo flow and other dynamic in-vivo phenomena, such as heart motion. Rotating Ultra-Fast Imaging Sequence (RUFIS) developed in this Center uses free induction decays (FID) for image reconstruction, and usually the data points at the begining of the FID are not available because of finite recovery times of the hardware. We improved image reconstruction by substituting a Linear Algebra Method (LAM) to reconstruct image profiles directly from FID data. Simulations confirm that the LAM method can accurately reconstruct 1 D projections from oversampled data even when 1/4 of k-space is missing or noise is present. Simulations were also performed to examine and confirm the minimal effects of turbulence and diffusion on signal loss. Using a spin echo-based velocity-encoding preparation sequence followed by RUFIS, quantitative velocity profiles of laminarand turbulent flow have been obtained. Flow has been quantified inside a straightglass phantom, emerging from a stenosis, and traveling through a curved channel. We also used RUFIS to analyze flow boundary reattachment and the effects of secondary flow patterns in curved flow. We also began to adapt RUFIS sequences for in vivo imaging of the rat. RUFIS can be used as an accurate non-invasive technique for distinguishing laminar and turbulent flow, measuring flow velocities, observing eddies, and identifying shear boundary reattachment downstream of a stenosis.
Showing the most recent 10 out of 148 publications
