Comparison of Newtonian and Non-newtonian Fluid Models in Blood Flow Simulation in Patients With Intracranial Arterial Stenosis

Background: Although blood is a shear-thinning non-Newtonian fluid, the Newtonian fluid assumption is most commonly applied in computational fluid dynamics (CFD) modeling in intracranial atherosclerotic stenosis (ICAS). We aimed to investigate the differences in cerebral hemodynamic metrics in ICAS, as quantified in CFD models with Newtonian and non-Newtonian fluid assumptions. Methods: We performed static CFD simulation with Newtonian and non-Newtonian (Casson and Carreau-Yasuda) fluid models, in a virtual artery model with an eccentric, 75% area stenosis. We then performed static, Newtonian and non-Newtonian CFD simulations in three patients with ICAS of different severities of area stenosis. We also performed transient, patient-specific, Newtonian and non-Newtonian simulations in a patient with ICAS. In each case, we compared the following hemodynamic metrics between CFD models derived with Newtonian, Casson and Carreau-Yasuda fluid assumptions: pressure distribution, translesional pressure ratio (PR, relative change in pressure across ICAS), and wall shear stress (WSS) distribution. Results: In all the static and transient simulations, the Newtonian/non-Newtonian differences on pressure distributions and PR values were negligible. In static CFD models of ICAS cases, there were more significant between-model differences in WSS distribution in low-WSS areas than in high-WSS areas. In the transient patient-specific CFD model of ICAS, there were considerable between-model differences in WSS distribution in low-WSS areas immediately distal to the stenotic region, especially during the diastolic phase. Conclusion: Newtonian fluid model may be applicable in pressure quantification in ICAS, but caution needs to be taken when using the Newtonian assumption in quantifying WSS, especially in low-WSS areas.