Patient-specific Computational Modelling of TAVI: Engineering Advances Toward Clinical Application

Transcatheter Aortic Valve Implantation (TAVI) has emerged as a pivotal therapeutic intervention for patients with severe aortic stenosis who are considered unsuitable for conventional open-heart surgery. Despite its clinical efficacy, pre-procedural planning remains challenging due to the complexity of patient-specific anatomies and the intricate biomechanics associated with device deployment, particularly in anatomically complex cases. Computational modelling has increasingly been recognized as a powerful approach to simulate TAVI procedures, anticipate procedural outcomes, and support individualized treatment planning. However, existing methodologies are constrained by significant limitations, including methodological heterogeneity, incomplete physiological representation, limited experimental and clinical validation, lack of standardization, and insufficient integration into routine clinical workflows. This doctoral research addresses these gaps through the development of a comprehensive, validated, and clinically interpretable computational framework for patient-specific TAVI simulations. The proposed framework integrates Finite Element Analysis, Computational Fluid Dynamics, and Fluid-Structure Interaction simulations, encompassing the entire workflow from robust image-based anatomical reconstruction to the derivation of clinically relevant outcome metrics. High-fidelity TAVI device models were generated via reverse engineering and calibrated through experimental testing to ensure accurate representation of mechanical behaviour. The incorporation of arterial pre-stress and patient-specific, time-varying boundary conditions further enhanced physiological fidelity. Additionally, a novel computational workflow was implemented to simulate balloon post-dilation, allowing systematic assessment of its effects on stent deformation and paravalvular leakage risk. Specialized workflows were developed to address complex clinical scenarios, including bicuspid aortic valves, TAVI-in-TAVI interventions, and the presence of mechanical mitral valves. Verification and validation of the computational framework were rigorously conducted in accordance with ASME V&V 40 guidelines, utilizing post-operative imaging and clinical hemodynamic data to validate the accuracy and reliability of structural and fluid dynamic predictions. Exploratory integration of machine learning techniques further demonstrated the potential to augment mechanistic simulations with rapid, data-driven predictive capabilities. Moreover, engineering outputs were translated into clinically interpretable indices, such as paravalvular leakage grading and risk of conduction disturbances, and communicated through standardized, interactive pre-procedural planning reports to facilitate clinical adoption. The framework developed herein significantly advances the predictive reliability, physiological realism, and clinical applicability of in-silico TAVI simulations. By enabling accurate replication of device–anatomy interactions and providing robust decision-support tools, this research contributes to improve pre-operative planning, optimize device selection and positioning, and mitigate procedure-related complications. Beyond TAVI, the principles and methodologies established in this work offer a blueprint for computational modelling in other structural heart interventions, thereby promoting the integration of in-silico models within precision cardiovascular medicine.