Abstract Body (Do not enter title and authors here): Background:
Percutaneous coronary intervention (PCI) of bifurcated coronary lesions is associated with higher complication and restenosis rates due to complex vessel geometry and plaque heterogeneity. Accurate preprocedural planning is essential, and computational simulations have emerged as valuable tools. Finite element analysis (FEA) allows detailed simulation of stent deployment and vessel biomechanics during PCI. Among FEA approaches, full-plaque (FP) models, which incorporate the lumen, vessel wall, and heterogeneous plaque components, are the most comprehensive. This study presents development and validation of a high-fidelity FP model using nine patient-specific coronary bifurcation cases.
Methods:
Pre-stenting OCT and angiographic data from nine patients with LAD bifurcation lesions were used to reconstruct 3D vessel geometries, including fibrotic, calcified, and fibrolipid plaques (Image 1). The FP model was meshed with tetrahedral elements. Plaques were assigned literature-based nonlinear elasto-plastic properties; vessel walls were modeled as hyperelastic. Stepwise PCI procedures were simulated using the Abaqus/Explicit solver. Simulated lumen geometries were validated using post-stenting OCT, and Bland-Altman analysis was used to assess agreement.
Results:
The FP model closely reproduced all procedural steps and biomechanical changes, including balloon and stent deformation, plaque displacement, and vessel remodeling (Image 2). Mean lumen diameters from the FP model showed strong agreement with post-stenting OCT (mean bias 0.07 mm; 95% limits of agreement: -0.38 to 0.51 mm) (Image 3a). It also replicated cross-sectional lumen shape and stent strut distribution, including regions of malapposition (Image 3b), key for identifying potential mechanical issues such as high stress areas or incomplete stent apposition that could impact long-term outcomes.
Conclusion:
This study presents the development and validation of a high-fidelity FP model for simulating coronary bifurcation stenting. Built from OCT-based segmentation and literature-derived material properties, the model accurately captures anatomical detail and biomechanical behavior. Its strong agreement with post-stenting OCT supports its use as a reference standard for evaluating simplified models. Future efforts will focus on automating the workflow to enable real-time clinical application.