Evaluating Endovascular Approaches: A Simulation Framework for Posterior Circulation Stroke Management

Introduction: The use of mechanical thrombectomy (MT) is now a standard treatment for posterior circulation strokes (PCS). However, the best approach for MT in basilar occlusions is not well established due to scarce data. This study uses computational models to assess how arterial access impacts navigation difficulty to the basilar artery and aims to determine the most efficient access route.

Methods: We compared navigation tortuosity in MT cases using both vessel centerlines from imaging and computationally simulated device navigation in patient anatomies. 23 patients were utilized to generate vascular models from CT angiograms. For each model, we generated vessel centerlines for navigating from a femoral approach to the distal left and right vertebral arteries and a right radial approach to the distal right vertebral artery. We also simulated a 0.035 wire navigating these 3 routes. Deviations between centerlines and simulated pathways were measured by normalizing the area between curves by the path length. For all three routes, local bending energy (LBE) and tortuosity index (TI) were calculated based on vessel centerlines and simulated pathways.

Results: Each anatomy yielded 3 pairs of LBEs and TIs. All paths exhibited some degree of variation between vessel centerline and device pathway, with femoral approaches yielding greater differences than radial approaches. Paired t-testing revealed that all simulated LBEs were significantly lower than the corresponding values from vessel centerlines. Device TI was significantly greater than its vessel centerline counterpart for the femoral artery to left vertebral artery route.

Conclusion : Our results demonstrate several key findings: 1) we have developed a framework to generate simulations of device navigation in subject-specific anatomical boundaries, and 2) the outputs of these simulations are significantly different from those computed using traditional vessel centerlines to evaluate TI. This supports our hypotheses regarding the need for mechanics-based considerations in evaluating patients' CTAs ahead of endovascular interventions to reduce navigation challenges. TI does not accurately model navigation challenges as devices will seek the lowest possible energy state when navigating, likely deviating from the centerline.