【作者】 付文宇；

【导师】 尚德广； 乔爱科；

【作者基本信息】 北京工业大学 ， 机械设计及理论， 2010， 博士

【摘要】 随着技术进步,支架介入治疗心脑血管动脉瘤越来越普及。支架结构、支架丝截面形状、通透率、支架放置位置、支架连接体形状及支架疲劳特性等这些因素会对支架治疗动脉瘤产生不同的影响。在相等通透率条件下,对不同结构、不同截面形状支架对个体化颈内动脉瘤治疗有何不同的血液动力学方面的影响,还没有这方面的研究报告。为了研究支架对个体化脑动脉瘤血液动力学特性的影响,首先必须建立基于医学扫描图像的个体化血管模型。本文基于逆向工程思想构建了人体个体化动脉瘤模型,进行了动脉瘤及其支架治疗的血液动力学数值模拟研究。以提高个体化动脉血管进行计算流体动力学分析效率和减小计算误差为评价指标,探讨了两种从医学扫描图像到数值模拟计算流程的优缺点(分别基于STL模型与NURBS模型)。以快速准确构建动脉血管表面模型为标准,探讨了不同图像分割方法的优缺点。以提高动脉血管模型体网格质量为目标,探讨了如何提高表面模型面网格质量。依据上述建模方法,分别构建了一个胸主动脉和一个颈内动脉瘤血管模型,以便于进行生物力学分析。在此基础上,按照逆向工程中生成NURBS曲面模型的方法,建立了颈内动脉瘤血管壁模型。所建立的个体化颈内动脉瘤模型,不仅具有一个侧壁动脉瘤,而且有一个较大的S型弯曲。为了研究Von Mises Stress在血管壁面上的分布特性及随时间变化特性,对建立的个体化颈内动脉血管模型进行了瞬态流固耦合计算。对建立的血管壁模型进行了静态有限元分析,以确定是否可用静态有限元分析替代流固耦合计算来推断Von Mises Stress分布特性。经过对比分析,发现虽然个体化血管模型几何形状复杂,但静态有限元分析获得的Von Mises Stress分布特性可以替代流固耦合计算,两者之间相差不大。根据临床实践测得的血流速度值同数值模拟获得的数据进行了对比分析,初步验证了数值模拟的准确性。为了探讨支架结构对动脉瘤血流动力学的影响,设计构建了五种不同结构、不同截面形状裸支架(通透率近似相等)的三维实体模型。利用计算流体力学方法,借助于有限元软件ANSYS CFX11.0对五种支架分别植入颈内动脉瘤前后进行数值模拟研究。提取和分析了支架植入动脉瘤前后瘤腔内的流线、速度等值面面积、壁面切应力和压力等血流动力学参数。结果表明支架植入动脉瘤后瘤腔内的血流速度被削弱,血流对动脉瘤远端冲击面积减小,其中以圆形截面螺旋支架及矩形截面网格支架冲击面积减小最多。植入支架的五个模型中动脉瘤顶部局部较高壁面切应力区域完全消失,动脉瘤壁面压力略有增加,但增加值可忽略不计。本文摸索出来的方法具有较强的通用性,不仅适用于人体动脉血管,同样可以用于建立类似的血管甚至其他组织的个体化有限元模型。依据医学扫描图像建立的个体化模型在几何形状上更接近真实情况,为真实反映血液真实流动情形(特别是局部流动特性)打下了一个坚实的基础。本文的主要创新点如下:(1)对具有较大S型弯曲且带有侧壁动脉瘤的个体化颈内动脉模型进行了瞬态流固数值模拟计算。经过对比数值模拟结果发现,在动脉瘤血流动力学数值模拟中,可以使用静态有限元分析替代流固耦合计算获得的内壁面Von Mises Stress大小和分布特性。(2)在近似相等通透率条件下,对五种不同支架对动脉瘤腔的血液动力学参数的影响进行了数值模拟计算研究,综合考虑动脉瘤远端血流冲击面积减小量、动脉瘤WSS减小情况及支架弯曲变形能力,发现矩形截面网格支架具有较优的性能。这可以为支架结构设计提供一些理论指导意义。(3)通过计算发现,在植入支架后,动脉瘤壁面压力值略有增加(同没有支架的血管模型比较)。这同已有的一些研究报告结论正好相反,这主要是由于使用的入口边界条件不同而产生的。本文使用的均为速度的入口边界条件,而有些研究人员使用的是压力入口边界条件。这两者之间的本质区别是,前者血管入口的流量是相同的(植入支架的模型同没有支架的模型相比),而后者血管入口的流量是不同的。通过对网格支架与螺旋型支架进行静力有限元分析,发现网格支架的柔顺性(包括弯曲变形能力与扭转变形能力)好于螺旋型支架。更多还原

【Abstract】 With the development of technology, it is very popular to use stent to treat vascular aneurysm. Many factors, such as stent structure, cross-section of stent wire, porosity, placement of stent, shape of stent connector and fatigue characteristics of stent will have different influence on stent intervention of aneurysms. It hasnot been reported about what different impacts on the hemodynamics of treatment for patient-specific internal carotid aneurysms will be made by stents with different structures and different cross section of strut at the same porosity. In order to perform the study mentioned above, first patient-specific vessel model must be established based on medical scan images. Based on the idea of reverse engineering, patient-specific aneurysimal model was constructed. Numerical simulation of hemodynamics was performed to the aneurysimal model and the models with stent.Taken the calculation efficiency of Computational Fluid Dynamics and reducing calculation errors for patient-specific vascular model as evaluation index, advantages and disadvantages about two calculatin flows from medical images to numerical simulation were discussed (based on STL model and NURBS model respectively). Taken the construction of surface model of arteries fast and accurately as evaluation indexes, advantages and disadvantages about different methods of image segmentation were discussed. To improve the quality of surface mesh for arterial vessel model, specific methods of improving the quality of the surface mesh were discussed.In order to facilitate the study of biomechanics, a patient-specific thoracic aortic model and a patient-specific internal carotid aneurysm model were constructed respectively based on the method mentioned above. In accordance with the method to generate NURBS surface model used in reverse engineering, a vessel wall model of internal carotid artery was established. The established patient-specific model of internal carotid arterial aneurysm has not only a sidewall aneurysm but also a large S-shaped bend. Vessels with S-shaped bend are very common in the human vascular system and worthy of study using the established model. Pulsatile Fluid-Structure Interaction simulation was performed using the patient-specific internal carotid arterial model in order to study the spacial distribution characteristics and temporal changing characteristics of Von Mises Stress on the vascular wall. Static finite element analysis was carried out using the vessel wall model of internal carotid artery so as to determine whether Fluid-Structure Interaction simulation can be replaced by static analysis or not when studying distribution characteristics of Von Mises Stress. Although the geometry shape of patient-specific vessel model is very complex, comparason analysis showed that Fluid-Structure Interaction simulation can be replaced by static analysis when studying distribution characteristics of Von Mises Stress. Difference between them is small. A comparison study between the blood flow velocities obtained from clinical practice and those extracted from the result of the numerical simulation was performed, and a preliminary accuracy of the numerical simulation was verified in this paper.Five virtual stents with different structures and wire cross-sections were designed for incorporation into the same patient-specific aneurysm model in order to investigate the influence of stent structure on the hemodynamics of aneurysm. Computational fluid dynamics simulations were performed using ANSYS CFX11.0 so as to study how these five types of stent affect the hemodynamic parameters before and after intervention. Hemodynamic parameters such as streamline in aneurysimal cavity, velocity isosurface, wall shear stress and wall pressure were extracted and analysed for models with and without stent. Numerical results demonstrated that the mean flow rate in the aneurismal cavity decreased. Impact area of blood flow to aneurysimal distal wall reduced the most in the model that used a stent with a rectangular wire cross-section and spiral stent with a circular wire cross-section. Local high wall shear stresses at the dome of the aneurysm disappeared completely in models with stent. The wall pressure on the aneurysm increased slightly after implantation of the stent in all five models, but the increased value is small.The method mentioned above has the advantages of generality. It can be applied to the establishment of not only human arterial vessel but also other similar vessel and even patient-specific finite element model of other organs. Patient-specific model based on medical scan images is very close to the true vessel shape in geometry. It laid a solid foundation for reflection of the true blood flow (especially local flow characteristics). The main innovations of this paper are as follows: (1) Pulsatile Fluid-Structure Interaction numerical simulation was performed to the patient-specific internal carotid artery model with a large S-bend and a sidewall aneurysm. Comparison of the numerical simulation results showed that Fluid-Structure Interaction calculation can be replaced by static analysis when assessing magnitude and distribution of Von Mises Stress on aneurismal inner wall in numerical simulation of hemodynamics of aneurysm. (2) Five different stents were implanted into internal carotid model and their porosity is approximately same. Comparison studies about their effects to the hemodynamics parameters of aneurismal cavity were performed. The results demonstrated that stent with rectangular wire cross-section has a better performance when comprehensively considering the reduction of impact area to distal aneurismal wall, the reduction of WSS on aneurismal wall and bending ability of stent. These results can provide theoretical guidance for the design of endovascular stent. (3) It is found that aneurismal wall pressure increased slightly after stent implantation (comparison with the model without stent). This result is not consistent with published literature which stated that aneurismal wall stress decreases slightly after stent implantation. The reason for this difference is that the boundary conditions in the models are different. Comparing models with and without stent, the drop in pressure is set the same in some pulbished literature. However, in the present study, the inlet velocity is set to be the same for all the models. The blood flow conditions used in models with and without stent in some published literature are not the same. The blood flow velocity used in the model with stent is less than that used in the model without stent. After stent implantation, blood flow resistance increases. Therefore, with the same pressure drop, the inlet velocity would be reduced, and the blood flow used in model with stent would also be reduced. For the convenience of comparability, boundary conditions of equal blood flow should be used. Through the static structure finite element analysis to the mesh stent and spiral stent, it is found that flexility of mesh stent is better than spiral stent (including bending and twisting deformation capacity).更多还原