Hemodynamic Indicators of Stroke Risk Due To Carotid Stenosis: Computational Fluid Dynamics Approach

Stroke remains a leading cause of morbidity and mortality worldwide, with carotid artery stenosis as a major contributor. Current risk assessment primarily relies on stenosis severity, yet many strokes occur in patients with moderate stenosis. This highlights the need for improved predictive metrics. This study aims to identify key hemodynamic metrics for developing multi-parametric indices to improve stroke risk assessment beyond stenosis severity. We performed computational fluid dynamics (CFD) analysis on patient-specific carotid artery models derived from computed tomography angiography (CTA) scans of 22 patients (12 stroke, 10 non-stroke). We extracted hemodynamic parameters, including wall shear stress (WSS), wall shear stress gradient (WSSG), velocity, vorticity, and pressure. We used the normalization techniques to standardize the values, and statistical analysis to evaluate their ability to differentiate stroke from non-stroke cases. While individual hemodynamic parameters showed limited discriminatory power, multi-parametric indices significantly differentiated stroke from non-stroke patients. Mean velocity and minimum pressure frequently appeared in the most differentiating multi-parametric indices, highlighting their relevance in stroke risk assessment. Among the multi-parametric indices, the Velocity-Pressure-WSS (VPW) index (p = 0.0362, two tail T-test) and Velocity-Pressure-Vorticity (VPV) index (p = 0.0470, two tail T-test) demonstrated the strongest discriminatory power. This study reveals the importance of assessment by multi-parametric hemodynamic indices over single-parameter evaluation and establishes a foundation for hemodynamic-based stroke risk assessment by identifying four key parameters—mean velocity, minimum pressure, WSS threshold (WSSthres), and vorticity threshold (Vortthres)—that effectively differentiate stroke from non-stroke cases when combined. The VPW and VPV indices highlight the potential of multi-parametric approaches to improve risk stratification, setting the stage for future research in personalized stroke prediction models.