Computer Modeling of Neurovascular Flow Diverter
Intracranial aneurysm rupture is one of the main contributors (15%) to subarachnoid stroke, which is identified as the third deadliest disease in North America. Among current widely used treatment strategies, flow diversion represents the most recent treatment paradigm shift. The use of the flow diverter (FD), a densely braided, stent-mesh device has achieved superior occlusion/cure rate in long term (6-12 months) clinical follow-ups for traditionally difficult-to-treat aneurysms (wide-necked, large or giant, and fusiform/dissecting aneurysms). However, current FD application is compromised by a 6-8% complication rate, including in-stent thrombosis, posttreatment aneurysm rupture, and parenchymal vessel hemorrhage etc. The highly flexible FD construct also causes technical issues during the deployment. Furthermore, the delayed aneurysm occlusion induced by flow diversion makes it difficult for clinicians to predict the treatment outcome. Recently, neurointerventionalists have been using the dynamic "push-pull" technique during the FD implantation to manipulate the FD mesh density for enhanced flow diversion. However, the clinical deployment results using this technique could not be evaluated due to the limited resolution of the angiogram, therefore hindering its future application. These difficulties and challenges underscore the need for comprehensive understanding of the procedure of FD deployment and the resulting 3D hemodynamics in patient-specific aneurysms. To this end, we developed a finite element analysis (FEA) based workflow, the high fidelity virtual stenting (HiFiVS) technique, to simulate the complete clinical processes of deploying the FD and provide accurate account for the final FD geometry. The developed HiFiVS was preliminarily validated using the mechanical testing data of a braided stent and the x-ray recording of the FD unsheathing procedure. Further in vitro validation was implemented, where FD samples were deployed into patient-specific aneurysm phantoms using the push-pull technique to generate FD configurations with various mesh densities (dense vs. loose). The experimental deployments were then recapitulated by the HiFiVS. The experiment and simulation results were compared qualitatively and quantitatively on FD's positioning and mesh configuration. Good agreement showed that the simulation accurately captured key operations of the push-pull technique observed in vitro. Image-based computational fluid dynamics (CFD) was performed based on the 3D geometries of virtually deployed FDs. The CFD results showed that FDs with higher mesh densities across the aneurysm orifice achieved enhanced aneurysmal inflow reduction than FDs with low mesh density. The dynamic push-pull technique was demonstrated to be effective in maximizing the flow diversion performance of the FD, and therefore was favorable for more immediate aneurysm occlusion. The HiFiVS has shown its unique capability to analyze different deployment strategies of the FD and accurately predict its final geometry. Combined with the CFD analysis, this modeling workflow presents a promising analytical tool for further optimization of the flow diversion treatment.