Biophysical and Statistical Modeling for Predicting the Progression and Regression of Cardiac Growth

Heart disease is the leading cause of death in the United States and affects roughly one third of the adult population. In particular, approximately one million Americans suffer from myocardial infarctions (MI) each year; as a result, they may experience maladaptive cardiac growth and remodeling that leads to heart failure (HF). In fact, around six million Americans currently suffer from HF. Many of these patients also experience electrical conduction delays that cause ventricular dyssynchrony and worsen HF. Heart failure is progressive, with continuing dilation of the heart contributing to ever-worsening pump performance; predicting the course of this remodeling and its modification by treatments could therefore provide important insight. Fortunately, biophysical and statistical modeling provides a low-risk, low-cost framework that allows us to not only gain a better understanding of post-MI remodeling but also predict the progression and regression of heart failure. Therefore, the goal of this work is to develop computational models that can predict the changes in mechanics, composition, and geometry that occur in the heart following MI and the ventricular remodeling that occurs in response to current therapies aimed at reversing HF. We utilize two different modeling approaches to predict the progression and regression of cardiac growth: (1) a biophysical mechanistic model of the infarcted left ventricle (LV) that predicts remodeling during post-infarction healing, and (2) a statistical modeling framework to predict ventricular remodeling and patient outcome following cardiac resynchronization therapy. Overall, the work in this dissertation explores the prevailing concept in biomechanics that the long-term remodeling of mechanically active biologic tissues such as the myocardium can be predicted based on regional mechanics, using two complementary approaches: biophysical models that explicitly link mechanics to remodeling, and statistical models that inform how much of the observed remodeling can be explained by mechanics.