Factors affecting the accuracy of pressure measurements in vascular stenoses from phase-contrast MRI

In this work the effects of noise, resolution, and velocity (flow) on the measurement of intravascular pressure from phase-contrast (PC) MRI are discussed. To elucidate these effects, we employed an axisymmetric geometry that enabled us to calculate pressures in <2 min on a Sun Ultra SPARC 10 workstation. To determine the effects of vascular stenoses, we fabricated several stenotic phantom geometries (with 50%, 75%, and 90% area stenoses), and performed both MRI and computational fluid dynamics (CFD) simulations for various flow rates for these phantom geometries. Noise with Gaussian statistics was added to the velocity field obtained from the CFD simulations. The pressure maps obtained directly from CFD simulations for our phantom geometries were compared with pressure maps derived by our algorithm when 1) the input was noise-corrupted velocity data from CFD, and 2) the input was PC-MRI data collected from the phantoms. The quantitative effects of noise, resolution, and flow rate on the accuracy of pressure measurements were determined. We found that for flow rates below the Reynolds number for turbulent flow, resolution is a more significant determinant of accuracy than SNR. Furthermore, if other parameters remain constant, increased flow rates may result in decreased accuracy.