Haemodynamics of stent-mounted neural interfaces in tapered and deformed blood vessels

Abstract

The endovascular neural interface provides an appealing minimally invasive alternative to invasive brain electrodes for recording and stimulation. However, stents placed in blood vessels have long been known to affect blood flow (haemodynamics) and lead to neointimal growth within the blood vessel. Both the stent elements (struts and electrodes) and blood vessel wall geometries can affect the mechanical environment on the blood vessel wall, which could lead to unfavourable vascular remodelling after stent placement. With increasing applications of stents and stent-like neural interfaces in venous blood vessels in the brain, it is necessary to understand how stents affect blood flow and tissue growth in veins. We explored the haemodynamics of a stent-mounted neural interface in a blood vessel model. Results indicated that blood vessel deformation and tapering caused a substantial change to the lumen geometry and the haemodynamics. The neointimal proliferation was evaluated in sheep implanted with an endovascular neural interface. Analysis showed a negative correlation with the mean Wall Shear Stress pattern. The results presented here indicate that the optimal stent oversizing ratio must be considered to minimise the haemodynamic impact of stenting.

Figure 1

The design pattern of the stent-based neural interface. The stent had a rectangular strut profile (70 μm width and 50 μm thickness). Twenty-four electrodes (D = 500 μm) were attached to the stent struts.

Figure 2

The workflow for deformed blood vessel wall model generation and CFD analysis. (a) A generalised blood vessel model with a stent-like neural interface. (b) The blood vessel was deformed by the expansion of the stent struts after the mechanical simulation. (c) Blood flow was simulated on the new deformed geometry to produce WSS results for the deformed model. (d) The deformed geometry is viewed from the outside and inside. (e) Tetrahedral mesh of the stent and blood vessel wall. (f) Blood vessel 3D segmentation from Micro-CT slices for sheep 2. (g) Blood vessel 3D coordinates extracted from Micro-CT slices for sheep 2. Stent artefacts were present, which made the reconstruction non-ideal for simulation.

Figure 3

CFD results for wall shear stress (WSS). (a) WSS contour of the blood vessel wall (the black arrow indicates the blood flow direction). (b) A magnified view of the WSS pattern around the stent and electrode. (c) Histogram showing the area of WSS < 0.5 Pa with various stent-to-vein ratios. The area of the stent is not included in the area percentage. (d) Axial WSS distribution along the length of the blood vessels.

Figure 4

CFD results for wall shear stress gradient (WSSG). (a) WSSG contour of the venous wall (the black arrow indicates the blood flow direction). (b) A magnified view of the WSSG pattern around the stent and electrode. (c) Histogram showing Area of WSSG > 200 Pa/m with various stent-to-vein ratios. The area of the stent is not included in the area percentage.

Figure 5

(a) Streamlines of blood flow over the stent struts and electrodes for Case 3 (deformation) and Case 5 (tapering). (b) Secondary flow magnitude at the blood vessel cross-section (dotted line) for Cases 2 and 3 without tapering. (c) Secondary flow magnitude at the blood vessel cross-section for Cases 4 and 5 with tapering.

Figure 6

Comparison to Experimental Results. Top—WSS contour of the stented sheep blood vessel. Bottom—Blue: Circumferential average WSS along the length of the simulated blood vessel, computed from the WSS contour above; the WSS axis was flipped to compare trends between WSS and tissue thickness. Orange: Stent-associated tissue thickness in three sheep measured using Micro-CT imaging. Dashed line: Mean WSS value in the sheep blood vessels without a stent inside. Inset: Correlation between WSS and tissue thickness.