A comparative CFD study of four inferior vena cava filters

Abstract

Computational fluid dynamics was used to simulate the flow of blood within an inferior vena cava (IVC) geometry model that was reconstructed from computed tomography images obtained from a real patient. The main novelty of the present work is that we simulated the implantation of 4 different filter models in this realistic IVC geometry. We considered different blood flow rates in the range between V =20 and V =80 cm /s, and all simulations were performed with both the Newtonian and a non-Newtonian model for the blood viscosity. We compared the hemodynamics performance of the different filter models, and we paid a special attention to the total drag force, F , exerted by the blood flow on the filter surface. This force is the sum of 2 contributions: the viscous skin friction force, which was found to be roughly proportional to the filter surface area, and the pressure force, which depended on the particular filter geometry design. The F force is relevant because it must be balanced by the total force exerted by the filter hooks/struts on the IVC wall at the attachment locations. For the highest V value investigated, the variation in F among filters was from 116 to 308 dyne. We also showed how the present results can be extrapolated to obtain good estimates of the drag forces if the blood viscosity levels change, ie, if the patient with a filter implanted is treated with anticoagulant therapy.

Hemodynamic Analysis of VenaTech Convertible Vena Cava Filter Using Computational Fluid Dynamics

“…Considering the low pulsation of the IVC blood flow and pressure, it is reasonable to assume the IVC blood flow is steady (Takizawa et al, 2012;Aycock et al, 2014Aycock et al, , 2016López et al, 2018). In theory, the blood flow over the VTCF is characterized by the viscous, incompressible Navier-Stokes (N-S) equations (Liu et al, 2018) as follows:…”

“…where γ is the local shear rate. The Carreau model has been widely used and validated as adequately accurate for characterizing the non-Newtonian properties of blood in many previous studies (Swaminathan et al, 2006;Fortuny et al, 2015;Aycock et al, 2016;López et al, 2018;Zhou et al, 2018). The model parameters are set as µ 8 = 0.00345 Pa•s (also the plasma viscosity), µ 0 = 0.056 Pa•s, λ = 3.313 s, and n = 0.3568.…”

“…At present, there is a consensus that an ideal IVC filter should be non-migratory, non-thrombogenic and if possible, while capturing clots efficiently and maintaining vena cava patency (Leask et al, 2001(Leask et al, , 2004Harlal et al, 2007;Lessne et al, 2016;Montgomery and Kaufman, 2016). Most of the aforementioned concerns depend, to a certain extent, on the hemodynamic characteristics of an IVC filter design (López et al, 2018), which can be directly observed in vivo by computed tomography and magnetic resonance imaging (Gaines et al, 2018) or investigated by some in vitro experimental techniques, such as photochromic flow visualization (Couch et al, 1997). However, both in vivo and in vitro measurements do not study certain pathological states or physical models with the use of IVC filters and are always limited to specific patients and the setup ability and by high cost (Swaminathan et al, 2006).…”

“…Comparatively, the numerical simulation based on in silico computational fluid dynamics (CFD) has better potential to be flexible regarding condition settings, reproducible and controllable regarding physical models, and relatively low cost (Swaminathan et al, 2006;Singer et al, 2009;Wang et al, 2019;Shar et al, 2020). Currently, CFD models are widely and successfully employed to simulate the blood flows around different designs of IVC filters (López et al, 2018). The results indicate that the stagnation and/or recirculation zone with low shear stress downstream from the filter possibly traps emboli for potential thrombogenesis (Singer et al, 2009).…”

“…The simulation data show fairly low levels of turbulence intensity downstream from the IVC filters that would not likely be responsible for platelet activation (Rahbar et al, 2011). CFD can also calculate the total drag exerted by the blood flow on the filter surface that needs to be balanced by the total force exerted by the filter hooks/struts on the IVC wall at the contact locations (López et al, 2018).…”