Physician Northwestern Memorial Hospital, United States
Purpose: Modeling blood as a perfectly Newtonian fluid is common in computational hemodynamics, but this simplification may not fully capture shear-thinning effects relevant under pulsatile flow near intravascular devices. Inferior vena cava (IVC) filters introduce disturbed flow where vorticity and shear forces may promote thrombosis or hemolysis. Despite the clinical importance, little work has compared Newtonian versus Carreau (non-Newtonian) fluid models under pulsatile flow near IVC filters.
Materials and Methods: Computational fluid dynamics simulations were performed for three commercially available IVC filters in a vessel model under pulsatile flow. Blood rheology was modeled as Newtonian or Carreau across hematocrit values 20–60%. Vorticity magnitude was quantified as a surrogate for microcirculatory stasis, a driver of filter-related thrombosis.
Results: Across all filters and hematocrits, Carreau modeling generally reduced vorticity compared to Newtonian flow, especially at lower hematocrits. For Filter #1, mean vorticity was significantly lower with Carreau at Hct 20% (0.221 vs. 0.286, p< 0.0001), 30% (0.156 vs. 0.202, p< 0.0001), and 40% (0.151 vs. 0.183, p=0.0002), but paradoxically higher at Hct 60% (0.111 vs. 0.086, p< 0.0001). Filter #2 showed similar trends with reductions at Hct 20–40% (all p< 0.01), with no difference at 50% (p=0.67) or 60% (p=1.0). Filter #3 again demonstrated lower Carreau vorticity at Hct 20–40% (all p< 0.001), but no difference at 50% (p=1.0) and higher vorticity with Carreau at 60% (0.116 vs. 0.139, p=0.0009).
Conclusion: Carreau modeling more consistently predicts reduced vorticity near IVC filters under pulsatile flow, especially at physiologic hematocrits (20–50%). At elevated hematocrits, the relationship reverses, suggesting complex interactions between viscosity and filter-induced turbulence. For interventional radiologists, these results emphasize the value of incorporating non-Newtonian models in device evaluation and design, as vorticity maps may identify regions prone to thrombosis or hemolysis. This approach may guide filter design, patient selection, and procedural decision-making to minimize complications.
