Accurate quantification of blood stasis in the left atrium (LA) and its appendage (LAA) is essential for understanding the mechanisms that underlie thrombus formation and stroke risk. Recent advances in computational fluid dynamics (CFD) enable patient-specific simulations of atrial blood flow, but these models depend critically on how cardiac wall motion is derived from clinical imaging. Since typical imaging modalities provide only O(10) frames per cardiac cycle, motion must be heavily interpolated to achieve the temporal resolution required by CFD solvers. In the first part of this seminar, we examine the sensitivity of LA flow simulations to the reconstruction of wall motion. Using high-resolution electromechanical (EM) models as a reference, we systematically reduce the temporal resolution and interpolate the wall kinematics, revealing how different reconstruction strategies influence global atrial flow patterns and, in particular, the residence time of blood within the LAA.

In the second part, we examine fibrosis and how its burden and spatial distribution alter LA mechanics and hemodynamics. We apply various fibrosis patterns to patient-specific EM models to generate spatio-temporal wall-motion fields, which are then used in CFD simulations. These results allow us to isolate the effects of fibrosis amount and topology on atrial motion and blood residence time. Whereas global LA stasis is primarily determined by fibrosis burden, LAA hemodynamics show a more complex sensitivity to the pattern and severity of remodeling.