Individualized Rupture Risk Assessment and Analysis of Mechanobiological Interactions in Abdominal Aortic Aneurysms

腹主动脉直径增加30 mm被称为腹主动脉瘤(AAA)，这种情况下危及生命的破裂风险随着直径的增加而增加。因此，已建立的指南建议对45-55 mm的腹主动脉瘤进行介入治疗。在这种情况下，必须考虑到只有1/3的AAA&GT；50 mm实际破裂，而破裂也可能出现在小得多的AAA。因此，精确、可靠和患者特有的破裂风险预测是非常有意义的。在临床上，破裂危险分层主要基于AAA直径标准，该标准并不十分具体。通常不包括对可能高度相关的参数，如形态、壁厚、血栓、钙化和主动脉壁强度的考虑。在这个项目的第一阶段，我们开发了一个个性化的破裂风险预测模型。它基于一个计算模型，该模型考虑了基于我们的患者队列的多元线性回归模型中患者的具体几何形状和材料属性。建立了一个包含80名患者和221个组织样本的数据库。对组织标本进行力学、组织学和免疫组织化学表征。非机械生物学替代模型已经发展起来，它可以预测不能非侵入性评估的AAA特性。这些信息进入数值模型以计算壁应力，与组织样本测试得出的壁应力相比，产生破裂风险的量化。研究表明，新的风险分层方法优于临床已建立的风险分层方法。在本项目期间，我们将通过重要的附加方面来增强破裂风险模型，即AAA时间级数和对模型预测质量的严格量化。此外，我们增加了数据库的大小，并增加了基于血清和样本的组织学和免疫组织学研究结果。体外实验将在原代培养的主动脉平滑肌细胞中进行，以确定人腹主动脉壁机械转导的功能关系。我们发展了一种新的机械生物生长和重塑(G&R)材料定律，该定律不仅考虑了力学状态，还考虑了AAA管壁退变的生物学方面。G&R模型参数将通过基于纵向医学图像数据集的逆分析过程来校准。高斯过程回归将应用于生物和力学参数，以确定具有统计学意义的替代模型参数，以纳入G&R材料配方和墙体强度预测。同时，高斯过程回归也将用于通过预测的G&R和材料属性值的可信区间来评估预测能力。最终的破裂风险分层模型将在一个小的患者队列中进行回顾性和前瞻性的评估。

A diameter increase of the abdominal aorta of >30mm is called an abdominal aortic aneurysm (AAA), where risk of a life threateningrupture increases with diameter. Established guidelines therefore recommend interventional therapy for AAA with 45-55mm. In thiscontext, it has to be considered that only 1/3 of AAA >50mm actually rupture, while rupture might also appear in much smaller AAA.Therefore, a refined, reliable and patient specific rupture risk prediction is of great interest. Clinically rupture risk stratification ismainly based on a AAA diameter criterion, which is not very specific. Consideration of potentially highly relevant parameters as e.g.morphology, wall thickness, thrombus, calcifications and aortic wall strength usually is not included. In the first period of this project, we developed a model for individualized rupture risk prediction. It is based on a computational model that considers patient specificgeometries and material properties from multilinear regression models based on our patient cohort. A data base of 80 patients and 221 tissue samples was created. Tissue samples were characterized mechanically, histologically and immunohistochemically. Amechanobiological surrogate model has been developed that predicts AAA properties that can not be assessed non-invasively. Theinformation enters the numerical models to compute wall stresses, which, compared to wall strength derived from tissue sample testing, yields a rupture risk quantification. It has been shown that the novel risk stratification can be superior to the clinically established. In this project period, we will enhance the rupture risk model by important additional aspects, namely the AAA progression in time and a rigorous quantification of prediction quality of the model. Additionally, we increase the data base in size and by additional serum and sample based, histological and immunohistological findings. In vitro experiments will be performed in primary aortal smooth muscle cells to identify functional relationships of mechanotransduction in human AAA wall. We develop a novel mechanobiological growth & remodeling (G&R) material law that not only considers mechanical state but also considers biological aspects of AAA wall degeneration. The G&R model parameters will be calibrated through an inverse analysis procedure based on longitudinal medical image data sets. Gaussian process regression will be applied to the biological and mechanical parameters to identify statistically significant surrogate model parameters to incorporate in the G&R material formulation andin wall strength prediction. Simultaneously, Gaussian process regression will also serve for the assessment of the predictivecapabilities by means of confidence intervals for the predicted G&R and material properties values. The final rupture risk stratification model will be evaluated retrospectively and prospectively on a small patient cohort.