arXiv:2602.21409v1 Announce Type: new
Abstract: The relationship between vascular morphology and hemodynamics is fundamental to understanding the natural history of cerebral aneurysms (CAs). While global geometric indices have been widely studied, the local interaction between luminal curvature and wall shear stress (WSS) remains poorly characterized. This study analyzed a large cohort of CAs to investigate how local surface morphology relates to hemodynamics. This was performed via CFD flow simulations of a set of 76 patient-specific CA geometries using the OpenFOAM library. Geometry and pulsatile inflow conditions were modeled based on patient arterial diameter and age. Blood was assumed to be a Newtonian incompressible fluid flowing in a laminar regime. We utilized a geometric framework to classify the aneurysm lumen into spherical-like and saddle-like patches based on Gaussian curvature. Our results demonstrate a robust, statistically significant correlation between these curvature types and hemodynamic metrics, regardless of rupture status or aneurysm type. Specifically, saddle-like patches, predominantly found at the aneurysm neck, are associated with high time-averaged WSS, a low oscillatory shear index, and intense near-wall vortical activity as identified by the lambda2-criterion. In contrast, spherical-like patches, dominant at the dome, correspond to regions of flow impingement characterized by lower time-averaged WSS and an elevated oscillatory shear index. These findings suggest that wall curvature is a primary determinant of local hemodynamics. By bridging the gap between local wall morphology and pathological wall markers, this work suggests that curvature-based mapping can serve as a powerful tool for identifying vulnerable regions susceptible to thinning and rupture. This objective geometric assessment offers valuable insights for risk stratification and the precision planning of endovascular interventions.