Het koppelen van microstructuur aan mechanica: Microschaalkarakterisering en modellering van arterieel weefsel

Cardiovascular diseases are the leading cause of death globally. When arteries are affected, like for atherosclerosis, their microstructure often undergoes complex remodeling processes such as collagen deposition, elastin degradation, leading to tissue stiffening. These changes significantly alter the mechanical environment of arteries and can compromise the effectiveness of mechanically driven treatments like balloon angioplasty or bypass surgery. Two essential proteins, elastin and collagen, are central to the passive mechanical behavior of arteries, and their remodeling is a hallmark of disease progression. Establishing a link between microstructural and mechanical properties is essential for a deeper understanding of their interrelationship. The first part of this thesis outlines its clinical and biomechanical context, emphasizing the limitations of current treatment strategies and the need for more precise imaging and mechanical data of arteries, as well as microstructure-informed models. It highlights the importance of reliable 3D imaging techniques and reliable image analysis protocols to extract meaningful, robust and consistent microstructural parameters. It demonstrates the need for novel computational method enabling to take directly into account the microstructure of arteries instead of phenomenological relationship, as microstructural changes are a hallmark of disease progression and treatment failure, understanding their influence on the mechanical behavior of arteries is essential. Novel approaches, such as representative volume element (RVE) modeling, enable a discrete representation of individual microstructural components. Compared to traditional tissue-scale constitutive models, they offer greater flexibility to account for variations in elastin and collagen volume fractions as well as fiber orientation. The main objective of this thesis is, therefore, to develop a robust, microstructure-inspired microscale model that integrates both mechanical and imaging data. The first objective of this work was to establish a reliable imaging and analysis protocol for characterizing arterial microstructure. Chapter 3 compared multi-photon microscopy (MPM) and contrast-enhanced computed tomography (CECT), including its cryogenic variant (cryo-CECT), on porcine and ovine arteries. Cryo-CECT demonstrated superior image sharpness compared to conventional CECT and enabled adventitial collagen visualization. MPM and (cryo-)CECT revealed consistent trends in elastin and collagen thickness, fiber straightness, and collagen fiber orientation. Furthermore, both modalities revealed a high degree of disorganization in the medial wall of ovine arteries, characterized by gaps lacking elastic lamellae. This finding raises concerns about the suitability of the sheep model in clinical research for studying cardiovascular treatments, as elastin fragmentation is often considered a hallmark of treatment failure. Chapter 4 introduced a novel enzymatic digestion protocol to selectively isolate and image elastin and collagen networks in 3D. This approach enabled, for the first time, large-scale visualization and quantification of the medial collagen network. The collagen fibers were found to be more axially aligned, and both elastin and collagen networks exhibited lower straightness in the axial direction compared to the circumferential direction. The second objective enabled mechanical characterization of arterial tissue. Chapter 5 evaluated several contrast-enhancing staining agents (CESAs) for their suitability in 4D-CECT (in situ mechanical testing with CECT imaging). Hafnium Wells-Dawson Polyoxometalate (Hf-WD POM) was identified as the optimal CESA, preserving native mechanical behavior while providing high-quality imaging. In contrast, polytungstate acid (PTA) and Lugol staining induced significant tissue shrinkage and increased stiffness. Furthermore, mechanical testing of enzymatically treated samples was performed to characterize their isolated mechanical properties in Chapter 7. Collagen-digested samples containing elastin, smooth muscle cells (SMCs), and the remaining ground matrix exhibited anisotropic behavior between the circumferential and axial directions. This observation aligns with the microstructural trends previously identified. It suggests that the elastin network in these samples is not isotropic, potentially due to differences in lamellae straightness between the two loading directions. Elastin-digested samples showed delayed collagen recruitment in the axial direction, also consistent with microstructural observations indicating lower straightness in that direction. The third objective was the development of a microstructure-inspired RVE model. Chapter 6 introduced a computational framework that discretely incorporates essential arterial components, such as elastic lamellae, elastin fibers, collagen bundles, SMCs, and ground matrix, based on literature-derived imaging data. The model enabled separate analysis of each component's contribution to the overall mechanical response and revealed limitations in classical constitutive models, such as the Gasser-Ogden-Holsapfel (GOH) model, which underestimated the roles of elastin and SMCs. A novel material formulation for elastic lamellae, based on a recruitment stretch principle, was introduced in Chapter 7 and outperformed traditional models in fitting experimental data. The model was calibrated using microstructural properties derived from enzymatically treated samples. To assess the robustness and sensitivity of the RVE model, a Sobol sensitivity analysis was performed using surrogate models. The analysis revealed that variations in microstructural parameters could lead to a fivefold difference in mechanical output, even within the range of healthy tissue values. This highlights the critical importance of detailed microstructural characterization and the need to account for biological variability and uncertainty in computational modeling. In conclusion, this thesis presents a comprehensive framework that bridges microscale imaging, mechanical testing, and computational modeling to advance our understanding of arterial biomechanics. It lays the groundwork for future studies on diseased tissues and provides a valuable tool for exploring hypothesis on cardiovascular disease progression.