The heart is a complex mechanical and electrical environment and small changes at the cellular and subcellular scale can have profound impacts at the tissue, organ, and organ system levels. The goal of this research is to better understand structure-function relationships at these cellular and subcellular levels of the cardiac environment. This improved understanding may prove increasingly important as medicine begins shifting toward engineered replacement tissues and organs. Specifically, we work towards this goal by presenting a framework to automatically create finite element models of cells based on optical images. This framework can be customized to model the effects of subcellular structure and organization on mechanical and electrophysiological properties at the cellular level and has the potential for extension to the tissue level and beyond.

In part one of this work, we present a novel algorithm is presented that can generate physiologically relevant distributions of myofibrils within adult cardiomyocytes from confocal microscopy images. This is achieved by modelling these distributions as directed acyclic graphs, assigning a cost to each node based on observations of cardiac structure and function, and determining to minimum-cost flow through the network. This resulting flow represents the optimal distribution of myofibrils within the cell. In part two, these generated geometries are used as inputs to a finite element model (FEM) to determine the role the myofibrillar organization plays in the axal and transverse mechanics of the whole cell. The cardiomyocytes are modeled as a composite of fiber trusses within an elastic solid matrix. The behavior of the model is validated by comparison to data from combined Atomic Force Microscopy (AFM) and Carbon Fiber manipulation. Recommendations for extending the FEM framework are also explored.

A secondary goal, discussed in part three of this work, is to make computational models and simulation tools more accessible to novice learners. Doing so allows active learning of complicated course materials to take place. Working towards this goal, we present CellSpark:​ a simulation tool developed for teaching cellular electrophysiology and modelling to undergraduate bioengineering students. We discuss the details of its implementation and implications for improved student learning outcomes when used as part of a discovery learning assignment.