E-030\u2005An instantaneous computer simulation tool for sizing flow diverters

Abstract

Introduction Flow diverters (FDs) can be difficult to size during the treatment of cerebral aneurysms (CAs). Complications such as malposition, perforator occlusion, prolapse, and migration are common. The recommended strategy for avoiding these complications is to select a FD size that will appose well to the vessel wall and cover at least 2–3 mm of the parent vessel.1 Here we present a mathematical model that can be used to appropriately size FDs in real-time. The tool was verified against 154 high-fidelity finite-element (FE) simulations of FD deployments and 17 clinical deployments. Methods A mathematical model that approximates FDs as mechanical springs was developed. The model uses the FD geometry (braiding angle, diameter, and length) and a direct path on the vessel centerline to predict post-deployment FD length. Post-deployment FD lengths predicted by the mathematical model were validated against 154 FE simulations of the Pipeline Flex (Medtronic, USA) FD in different CA vessels. The FE simulations are computationally expensive (6–24 hours for one simulation) because they consider the full geometry of the FD, delivery system, and clinical ‘push-pull’ deployment technique.2 The mathematical model was also compared to 17 clinical deployments that were reconstructed from flat-detector CT scans of CAs treated with the Pipeline FD. Results The table 1 below presents comparisons between the mathematical model and the FE and clinical deployments. Figure 1 presents a qualitative comparison of all three deployments.Abstract E-030 Table 1 Differences in FD length between the mathematical model, FE, and clinical deployments Mathematical model compared to: Mean absolute difference(mm) Standard deviation(mm) Length measurement uncertainty(mm) High-Fidelity FE Simulations (n=154) 1.62 1.45 0 Clinical Deployments (n=17) 3.54 1.85 0.37–1.56 Abstract E-030 Figure 1 FD deployment that was (a) reconstructed from clinical data, (b) simulated using high-fidelity FE, and (c) simulated using the mathematical model Conclusion The mathematical model predicted post-deployment FD lengths that were similar to FE results but at a drastically lower computational cost. Results also showed good agreement in device length between the mathematical model and clinical deployments. References Al-Mufti F, et al. ‘Bailout strategies and complications associated with the use of flow-diverting stents for treating intracranial aneurysms,’ in interv. Neuroradiol. doi:10.1159/000489016 Babiker B, Haithemet al, ‘Clinical validations of simulated neurovascular braided stent deployments’.In ann Biomed Eng. doi: 10.1136/neurintsurg-2018-SNIS.113 Disclosures H. Babiker: 4; C; EndoVantage. 5; C; EndoVantage. B. Chong: 4; C; EndoVantage. C. Baccin: None. F. Albuquerque: 4; C; EndoVantage.