ETD

• FEM
• left atrial appendage
• mesh morphing
• strain field
• thoracic aorta

Cardiovascular wall mechanical properties are closely associated with the onset and progression of severe diseases affecting the thoracic aorta and left atrium (LA), including aneurysm and cardioembolic stroke. Alterations in wall mechanics may lead to aortic rupture and influence tissue–device interaction, as in percutaneous left atrial appendage occlusion procedures in patients with atrial fibrillation. Despite their clinical relevance, current assessment mainly relies on echocardiographic imaging, which provides limited spatial and temporal resolution and is restricted to planar representations. Therefore, the estimation of in vivo deformation and strain fields from medical imaging may provide additional patient-specific information for improved risk stratification and treatment planning.
This study evaluates a non-rigid registration framework based on mesh morphing techniques to reconstruct displacement and strain fields from ECG-gated CT. The method was validated using FEM-derived deformations on idealized geometries to identify optimal operative conditions, and then applied to patient-specific models of the thoracic aorta and LA. Results show accurate reconstruction of deformation fields, with node-to-node errors below CT spatial resolution. Strain patterns are consistent with physiological behavior, while increasing geometric complexity leads to higher registration errors.
These findings suggest that non-invasive estimation of patient-specific mechanics may support identification of regions prone to disease progression and improve procedural planning.