Epshtein Mark, Shazeeb Mohammed Salman, Kühn Anna Luisa, Anagnostakou Vania, Raskett Christopher M, King Robert M, Goyal Mayank, Mendes Pereira Vitor, Arthur Adam S, Puri Ajit S, Fiorella David, Gounis Matthew J
New England Center for Stroke Research, Department of Radiology, University of Massachusetts Chan Medical School, Worcester, MA, USA.
Image Processing & Analysis Core (iPAC), Department of Radiology, University of Massachusetts Chan Medical School, Worcester, MA, USA.
Interv Neuroradiol. 2025 Apr;31(2):201-207. doi: 10.1177/15910199231158444. Epub 2023 Mar 6.
BackgroundNeurointerventionalists use in-vitro vascular models to train for worst-case scenarios and test new devices in a simulated use environment to predict clinical performance. According to the Food and Drug Administration (FDA), any neurovascular navigation device should be able to successfully navigate two 360-degree turns and two 180-degree turns at the distal portion of the anatomical model. Here, we present a device benchmarking vascular model that complies with FDA recommendations.MethodsOur vascular model was assembled from quantitative characterization of 49 patients who underwent CT angiography either for acute ischemic stroke caused by large vessel occlusion or for aneurysm treatment. Following complete characterization of these data, the vascular segments were 3D reconstructed from CT angiograms of 6 selected patients that presented with challenging anatomy. The curvature and total rotational angle were calculated for each segment and the anatomical parts that complied with FDA recommendations were fused together into a single in-vitro model.ResultsThe model was constructed containing two common carotid branches arising from a type two aortic arch and the dimensions of the overall model exceeded the recommendations of the FDA. Two experienced neurointerventionalists tested the model for navigation difficulty using several devices on an in-vitro perfusion system and concluded that the model provided a realistic, challenging scenario.ConclusionsThis model provides a first prototype designed according to FDA recommendations of cumulative angle while also integrating an aggregation of actual patient-specific anatomy. The availability of this clinically relevant benchmark model presents a potential standardized approach for neurovascular device testing.
背景
神经介入专家使用体外血管模型来针对最坏情况进行训练,并在模拟使用环境中测试新设备,以预测临床性能。根据美国食品药品监督管理局(FDA)的规定,任何神经血管导航设备都应能够在解剖模型的远端成功完成两个360度转弯和两个180度转弯。在此,我们展示了一种符合FDA建议的设备基准血管模型。
方法
我们的血管模型是通过对49例因大血管闭塞导致急性缺血性中风或接受动脉瘤治疗而进行CT血管造影的患者进行定量表征组装而成。在对这些数据进行全面表征之后,从6例具有挑战性解剖结构的选定患者的CT血管造影中对血管段进行3D重建。计算每个段的曲率和总旋转角度,并将符合FDA建议的解剖部分融合到一个单一的体外模型中。
结果
构建的模型包含从二型主动脉弓发出的两个颈总动脉分支,整体模型尺寸超过了FDA的建议。两名经验丰富的神经介入专家在体外灌注系统上使用几种设备测试了该模型的导航难度,并得出结论认为该模型提供了一个现实且具有挑战性的场景。
结论
该模型提供了第一个根据FDA关于累积角度的建议设计的原型,同时还整合了实际患者特定解剖结构的汇总。这种临床相关基准模型的可用性为神经血管设备测试提供了一种潜在的标准化方法。