Heart is a prime organ in the human body, and continuously adapting and evolving through growth and remodeling processes to maintain a balance between the demand and supply of blood and oxygen during physiological/developmental (i.e., from birth to adult) and pathological (i.e., various heart diseases) conditions. The exact mechanism of the progression of disease and the growth and remodeling processes are, however, unclear. While numerous experimental studies have been performed on animal models to investigate the mechanism of heart diseases, they are associated with some limitations. To address these limitations, computational frameworks based on idealized, and patient specific heart have been developed. Considering the short history of computational cardiac mechanics compared to experimental studies, many improvements are necessary to advance computational cardiac models. Here we developed both patient and animal specific computational models to investigate the mechanics found in 3 different heart diseases.First, we developed a computational growth framework based on human biventricular geometry to investigate the growth and remodeling processes associated with mechanical dyssynchrony, a disease caused by the asynchronous contraction of the left ventricle (LV). Cardiac mechanics was described using an active stress formulation and growth model was formulated based on volumetric growth framework. Through prescribing myofiber stretch as growth stimulus, our model can quantitatively reproduce the thickening and thinning of ventricular wall at the late and early activated regions, respectively, for two activation sites, namely, interventricular septum and LV free wall. The model is also able to reproduce global LV dilation found in mechanical dyssynchrony, which is consistent with reported experimental studies.Second, we developed a computational-experimental approach based on swine model of pressure overload to investigate the correlation between local growth as indexed by changes in regional thickness and local mechanical quantities. The LV pressure and volume data were acquired from 4 aortic constriction swine models to calibrate the model. From the analysis using the Pearson correlation coefficient, we found a strong correlation between local growth and local myofiber stress induced by an instant rise in peak systolic pressure due to aortic constriction.Third, we developed a computational framework based on idealized LV model to investigate how pathological features, such as a reduction in global longitudinal strain (GLS), myofiber disarray and hypertrophy, affects LV mechanics in hypertrophic cardiomyopathy (HCM), a genetic heart disease. In this modeling framework, LV mechanics was described using an active stress formulation and myofiber disarray was described using a structural tensor in the constitutive models. Both the LV function indexed by ejection fraction and stroke volume and mechanics indexed by circumferential and longitudinal strain were reduced with increasing myofiber disarray.Last, we developed patient specific computational models of LV using clinical measurements of 2 female HCM patients based on two different phenotypes (obstructive and non-obstructive) and a control subject. After calibrating our models with clinical data, the results showed that without consideration of myofiber disarray, peak myofiber tension was lowest in the obstructive HCM subject (60 kPa), followed by the non-obstructive subject (242 kPa) and the control subject (375 kPa). With increasing myofiber disarray, peak tension has to increase in the HCM models to match with the clinical measurements. The computational modeling workflow proposed here can be used in future studies with more clinical and experimental data.
Read
