Abstract
Atrial fibrillation (AF) is a supraventricular tachyarrhythmia characterized by un-coordinated atrial activation with consequent deterioration of mechanical function. Affecting an estimated 33 million people worldwide, AF is the most common ar-rhythmia and is associated with an increased long-term risk of other cardiovas-cular diseases. Personalized computational modeling provides a novel framework for integrating and interpreting the combined role of electrophysiology (EP) and biomechanics (BM) in the development and progression of AF.Personalized computational finite element models of the left atrium were generated using a statistics-based segmentation approach applied to high-resolution coronary computed tomography angiography data capturing the spatial heterogeneity of the myocardial wall thickness. The complex myofiber architecture in the left atrium was estimated using an automated algorithm based on local solutions of Laplace’s equation informed by and compared to anatomical and morphological images. The influence of a variable transmural microstructure on local activation times was quan-tified through EP simulations for all patients using individual transmural myofiber interpolation functions. Minor differences in the maximum local activation times within patients were observed suggesting a negligible effect. Biaxial mechanical ten-sion test data of human atrial tissue were reinterpreted using a microstructurally-based strain-energy function to inform BM inflation experiments. A spatial correla-tion between mechanical stress pattern and myocardial wall thickness was observed highlighting the importance of an accurate representations of the atrial geometry. Finally, an isolated Langendorff perfusion in the left atrium was simulated coupling EP and BM utilizing a biophysically-based contraction model. This model provides the first electromechanics (EM) simulation of left atrial contraction including the spatial heterogeneity of the myocardial wall thickness, the complex myofiber archi-tecture and model parameters for EP and BM derived from human data coupled with a biophysically-based active contraction model.
| Date of Award | 2018 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Steven Niederer (Supervisor) & Martin Bishop (Supervisor) |