A framework for the fluid structure simulation of ascending aortic aneurysms with inverse identification of wall material properties

The aorta, the body’s largest vessel, is prone to aneurysm due to high pressure, age, physiological and genetic factors. An aneurysm, a potentially fatal dilation and thinning of the wall, has rupture as the main risk. Predicting rupture is difficult since imaging and guidelines are often insufficient. Computational fluid dynamics (CFD) explores hemodynamics and disease progression but assumes rigid walls, limiting accuracy. Fluid–structure interaction (FSI), which accounts for vascular distensibility, allows more accurate analyses, with a more realistic simulation of the hemodynamics and estimation of structural stresses in the wall. This work applies a semi-automatic FSI pipeline to a patient-specific ascending thoracic aortic aneurysm (ATAA), including wall property identification for a more accurate biomechanical assessment. The aortic models at end-diastole and peak-systole were reconstructed from computed tomography angiography (CTA) images, while the patient’s heart rate, brachial pressure and peak transaortic (TA) jet velocity from echo-doppler were used as pipeline inputs. The solid domain was generated by extruding the lumen wall with a homogeneous 2 mm thickness, and the aortic wall was modelled as non-linear, isotropic, and hyperelastic with a Neo-Hookean model. Patient-specific mechanical wall vessel properties were estimated via an iterative inverse identification approach. A Young’s modulus within 0.8–3.2 MPa was assumed and used in a prestress simulation to derive the aortic wall pretensional state at end-diastole. Structural analysis then estimated AAo volume at peak-systole. The simulated systolic volume was then compared with the imaging-derived measurement. If the error between simulated and target AAo systolic volumes was below 1% the Young’s modulus was considered optimal, otherwise the initial range was bisected. For FSI boundary conditions, a flow rate curve from literature was scaled in amplitude using the TA velocity and in time using heart rate. At the outflows, three-element Windkessel models were used, with parameters fitted through 1D and 0D simulations to match patient brachial pressure. A two-way FSI simulation was performed using the arbitrary Lagrangian Eulerian (ALE) formulation. From results, wall shear stress (WSS) descriptors and structural stress/strain fields were extracted to assess the synergistic fluid-solid mechanics interaction and its potential role in aneurysm progression. A difference of -1.70% was assessed between the ascending aorta volume measured with imaging and the peak-systole obtained with FSI at the second cardiac cycle, demonstrating that this pipeline replicates in-vivo distensibility adequately. Time average wall-shear stress (TAWSS) in an interquartile range of 1.07-2.12 Pa was present through the artery with the highest values located at the mid ascending aorta, a potential rupture risk area. Maximum principal stress (MPS) at the peak-systole was present in an interquartile range of 59.11-84.46 kPa. A co-localization between hemodynamic and structural descriptors was made, trying to analyze the possible interaction between the blood flow and the aortic wall mechanics; in particular, the TAWSS–MPS analysis yielded a similarity index of 0.347. This thesis applied a personalized FSI pipeline to patient-specific thoracic aortic aneurysm geometries, with the goal of estimating aortic wall properties from clinically available data and providing a clearer understanding and more accurate characterization of aneurysm biomechanics.