Improved Methods for Agent-Based Modeling of Infarct Healing

Over one million people experience a heart attack each year in the US. The development of numerous treatments for acute MI, such as the use of thrombolytics and stenting, have pushed the survival rate of patients that experience MI to 95\%. Despite this high survival rate, post-infarction scar tissue decreases cardiac function, producing long-term health impacts in survivors. As a result, research has turned towards investigating the formation and maturation of infarct scar. However, one of the biggest obstacles to understanding and perturbing scar formation is the lack of relevant models. In vivo models are time-consuming, expensive, and often difficult to control between subjects making them less ideal for exploratory research. On the other hand, in vitro models which are easier to standardize and may require fewer resources, often fail to recapitulate the dynamics and complexity of infarct healing. Computational models, which are fast, inexpensive, highly controlled, and offer tunable complexity, are uniquely positioned to fill the gap in models of scar formation. Therefore, this research seeks to advance our understanding of scar formation and ability to assess treatments which address MI scar, by building an efficient and flexible computational model incorporating the most up to date experimental data on infarct scar collagen alignment. First, we will transfer an existing infarct ABM to an object-oriented programming platform and incorporate a discrete migration scheme (as opposed to the continuous scheme utilized in the existing model) to improve the computational efficiency and flexibility of the model. Second, I will develop a computationally efficient method of simulating diffusing signaling molecules compatible with an ABM framework. Third, I will match early collagen alignment of the ABM to experimental data by evaluating new and existing model features. Fourth, I will determine if modifying cell-cell interactions in the ABM can produce collagen alignment heterogeneity consistent with data.