Purpose: There is evidence that wall shear stress (WSS), or the frictional force of flowing blood on the vessel wall, is involved in the pathogenesis of aneurysms and vessel stenoses. The objective of this investigation was to create a framework for calculation of 2D WSS using velocity distributions extracted from high-speed angiography (HSA) sequences. The impact of different parameters on WSS estimation was explored.
Methods: A 3D-printed patient-specific basilar-aneurysm model was imaged under constant-flow conditions using 1000-fps HSA. The average/ maximum velocity values were determined pixelwise over the flow sequence to create velocity-mapping distributions. At each coordinate position along the vessel wall a normal vector was defined, along which the X/Y-components of velocity vectors were indexed with increasing distance from the wall. It was assumed that there is no flow through the vessel wall, i.e. the velocity at the wall was forced to zero. The shear rates were obtained from the spatial gradients of fitted curves, or the rate of change of tangential-velocity values along the direction of the normal. The effect of tangential-velocity curve fitting parameters was investigated.
Results: The proposed framework for calculating 2D WSS allows for manipulation of parameters such as the number and spacing of points used along the normal vector. Finer resolution of data points allows for truer estimation of velocity nearest the wall, the tradeoff being increased noise. Smoothing splines produced the best fit for tangential-velocity curves. The highest WSS values were found along the aneurysm dome, as well as the bifurcation region between the outflow vessels and aneurysm sac where the flow impinges on the wall.
Conclusion: A method to obtain average and peak WSS per pixel from in-vitro HSA sequences was described. The application of this method to HSA sequences has the potential to provide clinically useful information during a vascular intervention.
Funding Support, Disclosures, and Conflict of Interest: This research was supported in part by Canon Medical Systems Corporation and NIH Grant 1R01EB030092.