Development of Methodology to Estimate Fractional Flow Reserve Using Magnetic Resonance Imaging and Computational Fluid Dynamics
Hair, Jackson Brooks
MetadataShow full item record
Fractional flow reserve is the current gold standard for evaluating severity of coronary artery disease, but it is underutilized clinically due to its invasiveness. Recent efforts have worked toward developing non-invasive alternatives, wherein medical imaging data are used to construct patient-specific computational fluid dynamics models to simulate blood flow through the coronary arteries and calculate virtual fractional flow reserve. Magnetic resonance imaging is particularly well-suited for this application due to its ability to directly quantify both angiographic geometry and flow velocity. Therefore, the purpose of this thesis was the investigation and development of magnetic resonance techniques toward defining the patient-specific boundary conditions needed in computationally estimating fractional flow reserve. In Aim 1, we performed a series of computational simulations to determine what patient-specific flow information is needed to calculate virtual fractional flow reserve. Then, we tested phase-contrast magnetic resonance in a cohort of healthy volunteers to validate its ability to quantify coronary arterial flow. In Aim 2, several novel implementations of self-gated magnetic resonance angiography were investigated for their ability to characterize coronary arterial geometry. Tests were carried out in several cohorts of adult patients with congenital heart disease and a cohort of pigs to study the use of self-navigation and both four and five–dimensional golden-angle radial sparse parallel magnetic resonance. Optimizations of both the acquisition and reconstruction frameworks were explored. Altogether, these studies advanced the use of magnetic resonance angiography in interventional cardiology.