Preliminary Findings on the Correlation between Left Atrial Wall Shear Stress and Atrial Endocardial Voltage in Atrial Fibrillation Patients

Camilla Cortesi1, Matteo Falanga2, Jonas Leite3, Marie Shannon Soulez4, Nadjia Kachenoura5, Cristiana Corsi2
1Università di Bologna, 2University of Bologna, 3Sorbonne Universiter, 4Sorbonne Université, 5INSERM


Abstract

The purpose of this preliminary study was to investigate the possible association between left atrium (LA) wall shear stress, assessed with computational fluid dynamics (CFD), and areas with fibrosis, defined by low values of bipolar voltage (BV<0.5 mV), in patients with atrial fibrillation (AF). Analysis was performed in 3 AF patients. 4D flow MRI data and dynamic CT data were analyzed to derive the LA anatomical model and wall displacements from which two distinct dynamic CFD simulation domains were created. Patient-specific inflow boundary conditions at the pulmonary veins were extracted from the 4D flow MRI velocity field and applied to both models. Blood velocity field and time-average wall shear stress (TAWSS) were evaluated for the CT- and MRI-derived models over one cardiac cycle. The 3D electro-anatomical map of the LA, consisting of its complete anatomy and the BV map, was acquired for each patient during the AF ablation procedure and registered to the TAWSS map. When considering avarage values, the comparison of TAWSS maps with BV maps revealed an overall correlation between lower TAWSS values and lower voltage values (Figure, Patient 1). Conversely, higher TAWSS values are associated with higher voltage values in the LA (Figure, Patient 2). However, local differences in the spatial distribution of TAWSS values emerge between the MRI-derived and CT-derived models, affecting the correlation with BV values. These preliminary findings suggest that the computational domain has an effect on CFD simulation results, and the distribution of the simulated blood velocity fields should be evaluated considering the different anatomical models. We further aim at regionalizing the LA, for a more local and consistent evaluation of the associations between TWASS and BV values.