A computational workflow for modeling complex patient-specific coronary stenting cases

• Coronary stenting simulations require both validated stent models and reliable artery digital twins. • The proposed approach is suitable for modeling complex patient-specific coronary stenting cases. • Pre- and post-operative images are used for the model preparation and validation. • The clinical outcome of the recovered lumen area was satisfactorily matched by simulation results. • Simulations allow for the detection of malapposed stent struts. In the era of in silico clinical trials, it is of paramount importance to guarantee simulation reliability. In the field of coronary stenting, there is a need to couple validated stent models with credible digital twins of the arteries, whose mechanical behavior is commonly simplified to guarantee a balance between simulation complexity and computational time, namely usability. To this aim, the current work proposed a phenomenological approach suitable for the mechanical description of patient-specific coronary arteries undergoing coronary stenting in complex cases, e.g. bifurcations, exhibiting overstretching due to procedural choices. Pre- and post-operative images were used to prepare four vessel models and validate the outcome of multi-step structural stenting simulations in terms of recovered lumen area. Arteries were modeled improving a previous strategy by the authors, namely accounting for different mechanical properties in the media and adventitia layer, with an assigned hyperelastic response with a softening at higher strains to simulate the damage due to overstretching. Plaque components, which were identified from patient images, were classified into lipidic, calcified, and generic, and associated with different properties. The simulation results demonstrated a good match with the clinical outcome of all the stenting procedures, with errors lower than 15 % in terms of recovered lumen area. This proved the reliability of the proposed simulation framework improving the performances of the previous model, making it usable for interpreting also situations where the artery underwent overstretching. The proposed approach allowed to account for the in vivo conditions and have good performance when aiming at describing quantities such as lumen reopening and the presence of malapposed struts following stent deployment.