Abstract TP261: Clot Integration Factor for in-vitro Quantification of Stent-retriever Configuration Using Cone-beam CT

Abstract

Objective: Mechanical thrombectomy using stent-retriever technology provides acute ischemic stroke patients with beneficial treatment of emergent large vessel occlusion. Still, the disparity between recanalization rates and clinical outcomes calls for enhanced device designs and treatment strategies with better reperfusion efficiency. Intuitively, sufficient device integration with the embolus is pursued to maximize chances of first-pass success, thereby limiting clot fragmentation and intimal trauma during extensive clot manipulation. Our aim was to develop an imaging-based measure of device integration in reproducible phantom experiments that could help identifying differences in aspects of thrombectomy procedures that may be related to angiographic and clinical outcomes. Methods: Deployment variations of the Trevo(TM) ProVue/XP retriever were tested in an in-vitro model system that mimicked a MCA-occlusion with a silicone vascular replica and two visible clot models (soft elastic, and hard inelastic) with 8 trials/group. High-resolution cone-beam CT imaging was performed prior to retriever retraction. An image processing pipeline was devised that used level-set segmentation, path-tracing, and Steiner circumellipse fitting to extract and determine the interior volumes of the clot and the three visible wires. Clot Integration Factor (CIF) was calculated as the ratio of the volume of clot-device intersection to the clot volume. Results: Example of the device, its wires, and the clot: Average clot volumes were 90.4±12.2mm3. CIF was significantly different for two deployment variations when the device engaged a hard clot (difference between means, 95%CI=[0.003,0.187], p=0.043), but not a soft clot model (95%CI=[-0.152,0.263], p=0.567). Conclusion: In-vitro imaging-based quantification of clot integration can detect differences in deployment configuration relative to a clot, which may support procedural and design improvements.