Atrial Fibrillation (AF) is a common disease that significantly increases the risk of strokes. Oral anticoagulants represent the standard preventive treatment, but they involve severe drawbacks, including intracerebral bleedings. Since in patients affected by nonvalvular AF, the Left Atrial Appendage (LAA) is the primary source of thromboembolism, percutaneous closure of the LAA is a viable option for people unsuitable for long-term anticoagulant therapy. However, the complexities related to the implant procedure, occlusion devices and the anatomical variability hinder the pre-operative planning, resulting in unexpected outcomes. In this context, in-silico models may represent a powerful support tool providing clinicians with more detailed information. Nevertheless, few works focusing on numerical modeling of LAA occlusion devices have been presented so far, and a detailed process to assess the model credibility, verifying that different sources of uncertainty did not affect the prediction, is missing. This work aims to illustrate a process that allows to build and validate the numerical model of a commercial occlusion device starting from only one sample available and without data provided by the manufacturer. To better identify potential uncertainties, the validation followed a step-by-step process that led from individual device behavior assessment to interaction with deformable conduit evaluation.

心房颤动 (AF) 是一种常见疾病，会显着增加中风的风险。口服抗凝剂代表了标准的预防性治疗，但它们存在严重的缺点，包括脑内出血。由于在受非瓣膜性 AF 影响的患者中，左心耳 (LAA) 是血栓栓塞的主要来源，因此对于不适合长期抗凝治疗的患者，经皮封堵 LAA 是一种可行的选择。然而，与植入程序、封堵器和解剖变异相关的复杂性阻碍了术前计划，导致了意想不到的结果。在这种情况下，计算机内模型可能是一种强大的支持工具，可为临床医生提供更详细的信息。尽管如此，到目前为止，很少有专注于 LAA 封堵器数值建模的工作，并且缺少评估模型可信度的详细过程，以验证不同的不确定性来源不会影响预测。这项工作旨在说明一个过程，该过程允许从仅一个可用样本开始构建和验证商业封堵器的数值模型，而无需制造商提供数据。为了更好地识别潜在的不确定性，验证遵循一个循序渐进的过程，从单个设备行为评估到与可变形导管评估的交互。这项工作旨在说明一个过程，该过程允许从仅一个可用样本开始构建和验证商业封堵器的数值模型，而无需制造商提供数据。为了更好地识别潜在的不确定性，验证遵循一个循序渐进的过程，从单个设备行为评估到与可变形导管评估的交互。这项工作旨在说明一个过程，该过程允许从仅一个可用样本开始构建和验证商业封堵器的数值模型，而无需制造商提供数据。为了更好地识别潜在的不确定性，验证遵循一个循序渐进的过程，从单个设备行为评估到与可变形导管评估的交互。