Abstract

Background: Fluid dynamics play an important role in arterial diseases. PC-velocimetry as a magnetic resonance (MR) method is a useful tool for the examination of blood vessels, in particular in haemodynamics and morphology. Due to signal loss caused by turbulent flow, classification of stenoses remains difficult. In the presence of metallic stents, susceptibility effects and RF-artefacts further deteriorate image quality.

Materials and methods: Regarding fluid dynamics, stenosis, stent and restenosis (i.e., re-closing of the vessel in areas with stents) have similar properties: the flow close to an obstacle is accelerated and after a specific length, the so-called reattachment length, it redevelops into its original flow profile. Depending on the shape of the obstacle and its reattachment length, a method is developed to classify stenosis using MR-imaging. Enlarged, simplified stenosis models (phantoms) were examined and were also modelled using computational fluid dynamics (CFD).

Results: MR- and CFD-measurements show influences of the stenosis shape, especially the diameter. In graphical analysis (streamline and vector plots), a reattachment length of 6.7×step size could be determined under laminar conditions and nonpulsatile flow.

Discussion: Stenosis and other narrowing of vessels are classifiable more exactly using PC-velocimetry. From the flow behaviour in the area around the narrowing of the vessel wall, the character of the stenosis can be derived. But more parameters, as bended vessel shapes, pulsated flow, non-Newtonian flow behaviour as well as elastic properties of the vessels need to be incorporated.