Session: Rising Stars of Mechanical Engineering Celebration & Showcase

Paper Number: 148660

148660 - Micro and Mesoscale Models of Cardiac Tissue to Understand Cardiac Function 

Introduction
At the microscale, cardiac tissue consists of billions of cells, which form a complex network including frequent cell branching and cleavage planes. Studying the micro-mesoscale complex mechanics and how it leads to the observable macroscopic (e.g., at the ventricle level) motion is necessary to understand cardiac function, the causes of cardiac dysfunction, and which biomarkers can help in the early diagnosis of cardiac diseases.
The overall goal of this project is to link micro and mesoscale kinematics and mechanics to cardiac function and dysfunction. The first step consists in developing in silico models of cardiac tissue that reflect the complex microstructure observed experimentally and form the basis for subsequent mechanical and electrophysiology simulations. In the following, we describe the micro-mesoscale models' creation and their validation based on experimental data.

Methods
The model architecture focused on incorporating the realistic branching and interconnections of cardiomyocytes while maintaining extracellular volume fraction and anisotropy characteristic of the healthy myocardium. Initially, polygonal-shaped sheetlet regions (approximately six cells thick and fifteen cells wide) were defined in a representative cube with edge length equal to 500 microns. Subsequently, cardiomyocyte chains were created inside each sheetlet following a defined preferential direction perpendicular to the base of the polygonal cell. Along each chain, the cardiomyocyte orientation was allowed to deviate slightly from the preferential direction. Branches were added between chains at a frequency of approximately one branch every two cardiomyocytes. This estimated branching frequency was selected to reproduce the experimental observation that cardiomyocyte connectivity is in between a fully connected and a tract bundle network [1]. To validate the anisotropy of the constructed models, we simulated the diffusion of water molecules within the representative volume and compared the corresponding diffusion tensor with the ones reconstructed from previously acquired experimental data. Diffusion simulations were performed using 20,000 particles, varying time steps, and varying total duration to probe the water diffusion at different time scales and compare with experimental results obtained using different cardiac diffusion tensor imaging (cDTI) sequences. At the end of each diffusion simulation, the corresponding diffusion tensor was reconstructed and its eigenvalues and eigenvectors were compared to the measured experimental data.

Results
Different cardiac tissue models were generated with varying volume fractions, cleavage plane thickness, sheetlet shape, cardiomyocyte size, and tortuosity of the cardiomyocyte chains. The connectivity network of each model was extracted, and the largest connected cell network was plotted against the number of cell junctions. This connectivity plot was in between the fully connected and track-bundle connectivity data reported in the literature [1] and further confirmed by local preliminary human histological data. The diffusion simulations showed mean diffusivity and fractional anisotropy values consistent with values for healthy myocardium. In addition, the primary eigenvector of the simulated diffusion tensor is correctly aligned with the preferential direction of the cardiomyocytes. However, simulated secondary and tertiary eigenvalues are in the correct ratio only when higher diffusion times are probed, while their ratio is close to one for shorter diffusion times. This aspect is currently investigated as it may suggest additional micro and meso-structure features to be incorporated in the constructed models.

Conclusions
We have presented a method to generate in silico models of cardiac structure at the micro and mesoscale. The models have been validated using histological and cDTI data. Given the in silico nature of the models and the cell structure variability (e.g., cell branches, cell direction, sheetlets' width and configuration), many instances of these models can be generated to investigate the variability in kinematics and mechanics associated with different types of micro and meso-structures. Future work will focus on integrating the micro and meso-structure models we have developed in coupled electro-mechanical simulations of cardiac contraction and relaxation to study the relation between microstructure and mesoscale mechanics and kinematics.

References
[1] Wilson, A.J., Sands, G. and Ennis, D., 2021. Graph-Based Analysis of Cardiomyocyte Network Connectivity. Circulation, 144(Suppl_1), pp. A10881-A10881.

Presenting Author: Luigi Perotti University of Central Florida

Presenting Author Biography: Luigi Perotti received his Laurea (B.S./M.S.) degree in Civil Engineering from Politecnico di Milano, Italy, in 2004. Subsequently he continued his studies in Mechanical Engineering at Caltech where he received his M.S. in 2006 and his Ph.D. in 2011. At the end of 2011, he joined the Mechanical and Aerospace Engineering department at UCLA as a postdoctoral scholar where he pursued his research interests in biomechanics. In 2014 he joined the Radiological Sciences department at UCLA, first as a postdoctoral scholar and later (2016) as a project scientist. Dr. Perotti joined the Mechanical and Aerospace Engineering department at the University of Central Florida in 2019 where he leads the Computational Biomechanics Lab.

Authors:

Uditha Weerasinghage University of Central Florida
Kevin Moulin Boston Children's Hospital
Dennis Ogiermann Ruhr University Bochum
Daniel Balzani Ruhr University Bochum
Luigi Perotti University of Central Florida