A 3D Patient-Specific Model of Cerebral Blood Flow: Influence of Arterial Compliance and Circle of Willis Configuration

Abstract

We conducted an in silico study of blood flow in the brain using two different computational models: fluid–structure interaction (FSI) and conventional rigid wall (CFD). These models were applied to a patient-specific vascular network derived from MRI data. We used a mid-fidelity numerical approach called Transversally Enriched Pipe Element Method (TEPEM) to solve the governing equations. In the FSI model, we coupled the TEPEM strategy with an independent-ring model to account for arterial wall compliance. We compared the FSI and CFD models to understand how arterial wall distensibility affects pressure, flow, and the spatial distribution of flow-related properties. Additionally, we introduced three synthetic anatomical variations in the Circle of Willis to extend the comparison of the FSI and CFD models to these scenarios. Our results suggest that vessel compliance introduces discrepancies up to $2$ mmHg in distal cerebral regions and up to $15\%$ in the Wall Shear Stress. Regarding the anatomical variations on the Circle of Willis, the incomplete configuration introduces discrepancies in derived-flow quantities as the Time-Averaged Wall Shear Stress and the Relative Retention Time up to $20\%$.