Computational modeling may improve TAVR planning but clinical translation remains limited

Patient-specific digital simulation is emerging as a tool to support personalized planning of transcatheter aortic valve replacement (TAVR), particularly as the procedure expands to younger, lower-risk patients, and more complex anatomies. Despite procedural advances, complications such as paravalvular leak, conduction disturbances, coronary obstruction, and aortic injury remain important determinants of outcome. Current pre-procedural planning relies heavily on computed tomography-based anatomical assessment, which is indispensable but largely static and cannot fully capture dynamic device-tissue interactions, and haemodynamic mechanisms underlying many procedural events. Computational modelling derived from patient-specific imaging can extend this assessment by simulating valve deployment, device-tissue contact, and flow, offering mechanistic insight and potential support for individualized procedural decision-making. This systematic review evaluates modelling approaches addressing TAVR complications and procedural planning, including high-risk scenarios such as bicuspid valves and valve-in-valve procedures. Across the literature, modelling enables patient-specific simulations and exploration of procedural strategies that may reduce complication risk. However, clinical translation remains limited by small study populations, heterogeneous methodologies, limited patient-specific validation, and lack of integration into routine workflows. Future progress will require validation against clinically meaningful endpoints, scalable digital infrastructure, and close collaboration between clinicians and engineers to incorporate simulation outputs into routine Heart Team decision-making.