Assessing high-frequency fluctuations in intracranial aneurysms using direct numerical simulation, conventional computational fluid dynamics and particle tracking velocimetry*

An intracranial aneurysm (IA) is an artery expansion in the neurovascular system (NVS), carrying the risk of rupture, which can lead to subarachnoid hemorrhage. Despite the laminar flow behavior in the NVS, recent studies have detected high-frequency fluctuations (HFFs) in IAs. In this study three modalities were evaluated on their ability to capture HFF. Experimentally, particle tracking velocimetry (PT) and computationally direct numerical simulations (DNS) using an in-house solver and conventional computational fluid dynamics using STAR CCM+ (STAR) were applied. An aneurysm model was 3D printed for PT and a CT scan was captured, which was then processed for simulations. Three simulations/experiments were performed using three varying heart rates as inflow. Velocity and vorticity distribution was evaluated, and specific probes chosen for transient analysis. Results were analyzed to qualitatively and quantitatively validate the simulations with the PT and to compare the three modalities regarding their effectiveness to capture HFFs.

Evaluations showed a good agreement in time averaged and peak velocity and vorticity distribution between simulations and PT, which increased with voxelization of the simulation results. This indicates that detailed flow structures visible in the simulation results cannot be captured with PT. Transient probe analysis showed more similar trends between PT and simulation data in mid IA than in wall near areas, which might be an affect from the rigid, no slip wall setting in simulations. Comparing power spectrum at the analyzed probes in STAR and DNS, revealed tendencies to see HFFs in DNS increasing with heart rate. In conclusion, with simulations high resolved flow structures were captured better than with PT and with DNS HFFs can be found in the flow, not visible with STAR.