腹主动脉瘤的数值计算模型比较研究

目的分别采用纯流体模型和流固耦合模型来计算腹主动脉瘤的血流动力学特征,比较两种数值模型的不同,并讨论在研究腹主动脉瘤中的应用。方法使用Gambit 2.2.30和COMSOL Multiphysics 4.2建立腹主动脉瘤的理想模型,分别基于有限体的方法分析纯流体模型,基于任意拉格朗日-欧拉算法(Arbitrary Lagrangian-Eulerian)计算流固耦合模型。结果同样的入口速度下,纯流体模型出现4个涡流和6个局部压力集中;流固耦合模型只有2个涡流和局部压力集中,且涡流中心更接近腹主动脉瘤的远端。在边界层分离点、血流回帖位置以及腹主动脉瘤的近端和远端,两种模型均出现壁剪切力极值。血管壁的最大形变和最大壁应力出现在腹主动脉瘤的近端和远端。结论两种模型的涡流个数和涡流中心的位置均不一样,与瘤体的生长有着密切的关联;流固耦合模型中的最大壁剪切力比纯流体模型要小36%;最大壁应力和最大血管壁的形变量与出口血压呈正相关。在研究血管瘤生长与血流动力学的关系时需要考虑使用流固耦合模型。     

腹主动脉瘤几何形态对血液动力学影响的三维数值分析

目的研究腹主动脉瘤不同形态学对瘤内血液动力学的影响,为临床预估动脉瘤的破裂提供参考。方法根据动脉瘤影像学上的特点建立不同几何形态的数学模型,采用计算流体动力学(CFD)方法,在周期性脉动速度入流、刚性壁面以及血液为牛顿流体的条件下,对一个心动周期内瘤内流场进行数值分析研究,比较不同几何形态腹主动脉瘤内血液动力学。结果非轴对称模型可造成相对较大的壁面剪应力;带有峰值偏移和曲率半径偏转的腹主动脉瘤,瘤内漩涡的发展变化会随着几何形态的不同而产生变化。结论腹主动脉瘤内流场特征的变化受到不同形态学的影响。     

基于MRA的个体化腹主动脉瘤计算机仿真

目的探讨基于MRA图像进行个体化腹主动脉瘤(abdominal aortic aneurysm,AAA)计算机仿真研究的可行性,并从血流动力学层面探讨AAA的发生、发展和破裂机制。方法基于AAA患者的MRA数据采用逆向建模法建立AAA的三维几何模型;采用FLUENT软件进行数值模拟,假设血管壁为刚性壁,血液为不可压缩牛顿流体,建立瞬态模型。将收敛之后的数据导入到CFD-Post中进行结果分析,输出心动周期内不同时刻的血流流线图、流速分布图、血管壁面切应力分布图以及压力分布图。结果AAA瘤颈处血液流动的方式以层流为主,瘤腔内血流以涡流、湍流为主,且在瘤体膨大处较明显;瘤颈处血液流速快于瘤腔,瘤腔大部分区域在整个心动周期内都处于较低的流速水平,且波动不明显,瘤腔内的高流速区域多位于入口血流直接延续的部位;射血期的壁面切应力的量值及其变化幅度均大于充盈期,壁面切应力较高的区域总是分布于瘤颈附近,瘤腔的切应力在整个心动周期内始终处于较低水平;瘤体的壁面压力量值及其分布范围在射血峰值(t=0.08 s)时最大。加速射血期的壁面压力及其变化范围均较减速射血期及充盈期大。结论基于MRA图像可建立个体化的AAA计算机仿真模型,通过计算机仿真得到的AAA内血流分布规律对AAA的研究和临床个体化的诊治有一定的帮助。     

多种介入技术联合腔内修复术治疗复杂瘤颈腹主动脉瘤临床研究

目的:评估全腔内治疗复杂瘤颈腹主动脉瘤的安全性及早中期疗效。方法:回顾性分析2012年1月～2018年10月在柳州市工人医院血管外科诊治的腹主动脉瘤患者的临床资料。将符合复杂瘤颈并行全腔内治疗、临床资料完善的腹主动脉瘤患者纳入本研究。结果:共33例复杂瘤颈腹主动脉瘤患者进行了全腔内治疗,在进行传统的腔内修复术的同时,9例联合纤维蛋白粘合剂瘤腔内注射术,12例联合近端瘤颈Cuff支架植入,8例联合肾动脉支架植入,4例联合下行导丝牵拉技术。所有病例均获得技术性成功。在3个月～7年的随访过程中,3例术后出现1a型内漏,2例随访中瘤腔未见明显增大,1例再次行介入治疗后内漏消失。另30例患者在随访中未见腹主动脉瘤破裂、支架内漏、移位、感染等并发症。结论:根据瘤颈解剖特点,多种介入技术联合腔内修复术治疗复杂瘤颈腹主动脉瘤是安全可行的,早中期的随访提示疗效良好。     

腹主动脉瘤中定常流动的三维数值模拟

目的讨论定常流动情况下，三维腹主动脉瘤模型内的流动情况。方法应用三维非对称模型进行数值模拟。结果结果表明，在对称面上有一个涡而横截面上有两对涡的存在。壁面切应力在动脉瘤的出口处数值较高且变化大。结论流动和壁面切应力分布反应动脉瘤出口处为破裂的危险区域。     

二维弹性动脉瘤模型的血液动力学数值模拟与分析

本研究的目的在于探讨血液动力学参数对于动脉瘤形成、生长以及破裂的影响。从医学影像学图像出发,进行二维弹性动脉瘤模型的血液动力学数值模拟。将其结果与刚性动脉瘤结果在速度矢量场和切应力分布上进行对比,发现两者存在较大差异。主要表现在几个截面的速度分布有较大差异,其中一个出口的速度分布明显偏心,这将影响壁面的应力分布以及管壁物质交换。弹性和刚性模拟沿瘤壁的壁面切应力分布曲线走势基本相似,所不同的是两者的切应力数值特别是在瘤颈附近存在较大差异。通过分析阐明弹性壁模型更加符合临床与病理生理实际。研究结果对于分析动脉瘤形成、生长破裂以及预后有重要的临床应用价值。     

基于改进SymRG算法的腹主动脉瘤分割及自动定位方法

腹主动脉瘤是腹主动脉的局限性退化扩张,是一种高危致死性疾病。CTA是临床诊断腹主动脉疾病的首选方式。腹主动脉瘤的准确快速提取和定位对于腹主动脉瘤的诊断、预后评估以及手术计划的制定有重要意义。因此,本文提出了一种能够准确有效地实现腹主动脉内腔分割和瘤区层片自动定位的算法。首先,对CTA图像进行滤波预处理;其次,应用改进的快速对称区域生长算法,分别对单像素和合并的一维和二维区域进行判别和生长,标定连通区域;然后,用种子点提取目标三维连通区域,即为腹主动脉内腔;最后,计算腹主动脉内腔的横截面面积分布,自动定位瘤体位置。通过4例由GE64排容积螺旋CT扫描采集的胸部到左右髂动脉的CTA数据验证,本算法能实现腹主动脉内腔的准确分割和瘤区层片自动定位,自动分割结果和手动分割结果重合率达到95%以上,自动定位瘤区层片与手工标定的瘤区层片重合率达到96%以上。本文中提出的算法能够准确有效实现腹主动脉内腔的分割和瘤区层片的自动定位。     

腹主动脉瘤计算机仿真研究进展

腹主动脉瘤(abdominal aortic aneurysm,AAA)是腹主动脉较常见的疾病,破裂之后致死率极高,因此,及时发现和评估AAA破裂的风险就具有重要意义。近年来随着医学影像、计算机及血流动力学等理论和技术的快速发展,利用计算机模拟技术对AAA进行仿真研究已成为研究的热点,其模拟的结果趋于人体真实的动脉瘤,并在揭示AAA的发生及演变的机制方面发挥了重要作用。本文介绍了AAA仿真研究的原理及模型分类,详细地阐述了瘤体形态、瘤壁的结构及属性、血液及血流的属性、腔内血栓等相关因素在仿真研究中的作用,并对其目前的进展及局限性予以综述。     

组配式人工半骨盆假体重建后耻骨板断裂的非线性有限元分析

背景:目前有限元法应用于骨科相关生物力学分析主要是静态载荷下的研究,评价系统应力峰值及全场应力分布情况。但由于疲劳计算和体内力学环境复杂,很难明确临床相关的生物力学问题。目的:利用步行周期动态载荷分析组配式半骨盆假体重建髋臼缺损后耻骨板断裂的生物力学基础。方法:通过骨盆环三维有限元模型模拟骨盆Ⅱ+Ⅲ区肿瘤切除后骨缺损,并通过CAD软件绘制的半骨盆假体模型进行重建。重建后的三维有限元模型通过Abaqus6.7有限元分析软件分别计算重建骨盆在站立位及步行周期下耻骨板的应力分布情况,明确应力集中位置及耻骨板应力峰值。结果与结论:有限元计算结果收敛,站立位时耻骨板应力峰值为3.5MPa,应力集中点位于耻骨板近中点处;步行周期下耻骨板的应力峰值为273.0MPa,应力集中点位于耻骨板与人工髋臼连接处,与临床观察到断裂的部位一致。提示组配式半骨盆假体耻骨板断裂的原因主要是由于步行周期载荷下耻骨板应力集中部位引起的高周疲劳断裂。     

国人头颅三维有限元模型有效性检验

目的检验国人头颅三维有限元模型有效性。方法依据Nahum头颅冲击尸体实验参数，对有限元模型加载额部冲击载荷，载荷位于正中矢状面上，由前向后，呈正弦波形，峰值6．8kN，时程15ms。计算分析头颅结构的力学响应。分析不同部位结构的应力改变。由应力值计算颅内压，对模拟计算结果与尸体实验测定结果进行比较。结果头颅不同结构的应力改变不同，由额、顶及枕部硬膜节点主应力计算出的颅内压时间曲线与尸体实验测定的颅内压曲线吻合。结论由硬脑膜节点应力计算所得的颅内压与尸体实验测定值吻合。该头颅三维有限元模型可用于头颅冲击的模拟计算，模型有效。     

Towards A Noninvasive Method for Determination of Patient-Specific Wall Strength Distribution in Abdominal Aortic Aneurysms

The spatial distributions of both wall stress and wall strength are required to accurately evaluate the rupture potential for an individual abdominal aortic aneurysm (AAA). The purpose of this study was to develop a statistical model to non-invasively estimate the distribution of AAA wall strength. Seven parameters–namely age, gender, family history of AAA, smoking status, AAA size, local diameter, and local intraluminal thrombus (ILT) thickness–were either directly measured or recorded from the patients hospital chart. Wall strength values corresponding to these predictor variables were calculated from the tensile testing of surgically procured AAA wall specimens. Backwards–stepwise regression techniques were used to identify and eliminate insignificant predictors for wall strength. Linear mixed-effects modeling was used to derive a final statistical model for AAA wall strength, from which 95% confidence intervals on the model parameters were formed. The final statistical model for AAA wall strength consisted of the following variables: sex, family history, ILT thickness, and normalized transverse diameter. Demonstrative application of the model revealed a unique, complex wall strength distribution, with strength values ranging from 56 N/cm² to 133 N/cm². A four-parameter statistical model for the noninvasive estimation of patient-specific AAA wall strength distribution has been successfully developed. The currently developed model represents a first attempt towards the noninvasive assessment of AAA wall strength. Coupling this model with our stress analysis technique may provide a more accurate means to estimate patient-specific rupture potential of AAA.     

Effects of blood flow and vessel geometry on wall stress and rupture risk of abdominal aortic aneurysms

Sudden rupture of abdominal aortic aneurysm (AAA), often without prior medical warning, is the 13th leading cause of mortality in the US. The local rupture is triggered when the elusive maximum local wall stress exceeds the patient's yield stress. Employing a validated fluid-structure interaction code, the coupled blood flow and AAA wall dynamics were simulated and analysed for two representative asymmetric AAAs with different neck angles and iliac bifurcations. It turned out that the AAA morphology plays an important role in wall deformation and stress distribution, and hence possible rupture. The neck angle substantially impacts flow fields. A large neck angle may cause strong irregular vortices in the AAA cavity and may influence the wall stress distribution remarkably. The rupture risk of lateral asymmetric AAAs is higher than for the anterior-posterior asymmetric types. The most likely rupture site is located near the anterior distal side for the anterior-posterior asymmetric AAA and the left distal side in the lateral asymmetric AAA.     

Effects of blood flow and vessel geometry on wall stress and rupture risk of abdominal aortic aneurysms

Sudden rupture of abdominal aortic aneurysm (AAA), often without prior medical warning, is the 13th leading cause of mortality in the US. The local rupture is triggered when the elusive maximum local wall stress exceeds the patient's yield stress. Employing a validated fluid – structure interaction code, the coupled blood flow and AAA wall dynamics were simulated and analysed for two representative asymmetric AAAs with different neck angles and iliac bifurcations. It turned out that the AAA morphology plays an important role in wall deformation and stress distribution, and hence possible rupture. The neck angle substantially impacts flow fields. A large neck angle may cause strong irregular vortices in the AAA cavity and may influence the wall stress distribution remarkably. The rupture risk of lateral asymmetric AAAs is higher than for the anterior – posterior asymmetric types. The most likely rupture site is located near the anterior distal side for the anterior – posterior asymmetric AAA and the left distal side in the lateral asymmetric AAA.     

The effect of angulation in abdominal aortic aneurysms: fluid–structure interaction simulations of idealized geometries

Abdominal aortic aneurysm (AAA) represents a degenerative disease process of the abdominal aorta that results in dilation and permanent remodeling of the arterial wall. A fluid structure interaction (FSI) parametric study was conducted to evaluate the progression of aneurysmal disease and its possible implications on risk of rupture. Two parametric studies were conducted using (i) the iliac bifurcation angle and (ii) the AAA neck angulation. Idealized streamlined AAA geometries were employed. The simulations were carried out using both isotropic and anisotropic wall material models. The parameters were based on CT scans measurements obtained from a population of patients. The results indicate that the peak wall stresses increased with increasing iliac and neck inlet angles. Wall shear stress (WSS) and fluid pressure were analyzed and correlated with the wall stresses for both sets of studies. An adaptation response of a temporary reduction of the peak wall stresses seem to correlate to a certain extent with increasing iliac angles. For the neck angulation studies it appears that a breakdown from symmetric vortices at the AAA inlet into a single larger vortex significantly increases the wall stress. Our parametric FSI study demonstrates the adaptation response during aneurysmal disease progression and its possible effects on the AAA risk of rupture. This dependence on geometric parameters of the AAA can be used as an additional diagnostic tool to help clinicians reach informed decisions in establishing whether a risky surgical intervention is warranted.     

A patient-specific computational model of fluid-structure interaction in abdominal aortic aneurysms

It is generally believed that knowledge of the wall stress distribution could help to find better rupture risk predictors of abdominal aortic aneurysms (AAAs). Although AAA wall stress results from combined action between blood, wall and intraluminal thrombus, previously published models for patient-specific assessment of the wall stress predominantly did not include fluid-dynamic effects. In order to facilitate the incorporation of fluid–structure interaction in the assessment of AAA wall stress, in this paper, a method for generating patient-specific hexahedral finite element meshes of the AAA lumen and wall is presented. The applicability of the meshes is illustrated by simulations of the wall stress, blood velocity distribution and wall shear stress in a characteristic AAA. The presented method yields a flexible, semi-automated approach for generating patient-specific hexahedral meshes of the AAA lumen and wall with predefined element distributions. The combined fluid/solid mesh allows for simulations of AAA blood dynamics and AAA wall mechanics and the interaction between the two. The mechanical quantities computed in these simulations need to be validated in a clinical setting, after which they could be included in clinical trials in search of risk factors for AAA rupture.     

A Numerical Implementation to Predict Residual Strains from the Homogeneous Stress Hypothesis with Application to Abdominal Aortic Aneurysms

Wall stress analysis of abdominal aortic aneurysm (AAA) is a promising method of identifying AAAs at high risk of rupture. However, neglecting residual strains (RS) in the load-free configuration of patient-specific finite element analysis models is a sever limitation that strongly affects the computed wall stresses. Although several methods for including RS have been proposed, they cannot be directly applied to patient-specific AAA simulations. RS in the AAA wall are predicted through volumetric tissue growth that aims at satisfying the homogeneous stress hypothesis at mean arterial pressure load. Tissue growth is interpolated linearly across the wall thickness and aneurysm tissues are described by isotropic constitutive formulations. The total deformation is multiplicatively split into elastic and growth contributions, and a staggered schema is used to solve the field variables. The algorithm is validated qualitatively at a cylindrical artery model and then applied to patient-specific AAAs (n = 5). The induced RS state is fully three-dimensional and in qualitative agreement with experimental observations, i.e., wall strips that were excised from the load-free wall showed stress-releasing-deformations that are typically seen in laboratory experiments. Compared to RS-free simulations, the proposed algorithm reduced the von Mises stress gradient across the wall by a tenfold. Accounting for RS leads to homogenized wall stresses, which apart from reducing the peak wall stress (PWS) also shifted its location in some cases. The present study demonstrated that the homogeneous stress hypothesis can be effectively used to predict RS in the load-free configuration of the vascular wall. The proposed algorithm leads to a fast and robust prediction of RS, which is fully capable for a patient-specific AAA rupture risk assessment. Neglecting RS leads to non-realistic wall stress values that severely overestimate the PWS.     

A Pull-Back Algorithm to Determine the Unloaded Vascular Geometry in Anisotropic Hyperelastic AAA Passive Mechanics

Biomechanical studies on abdominal aortic aneurysms (AAA) seek to provide for better decision criteria to undergo surgical intervention for AAA repair. More accurate results can be obtained by using appropriate material models for the tissues along with accurate geometric models and more realistic boundary conditions for the lesion. However, patient-specific AAA models are generated from gated medical images in which the artery is under pressure. Therefore, identification of the AAA zero pressure geometry would allow for a more realistic estimate of the aneurysmal wall mechanics. This study proposes a novel iterative algorithm to find the zero pressure geometry of patient-specific AAA models. The methodology allows considering the anisotropic hyperelastic behavior of the aortic wall, its thickness and accounts for the presence of the intraluminal thrombus. Results on 12 patient-specific AAA geometric models indicate that the procedure is computational tractable and efficient, and preserves the global volume of the model. In addition, a comparison of the peak wall stress computed with the zero pressure and CT-based geometries during systole indicates that computations using CT-based geometric models underestimate the peak wall stress by 59 ± 64 and 47 ± 64 kPa for the isotropic and anisotropic material models of the arterial wall, respectively.