An Improved Computational Model Based on Cardiac Imaging Data for the Quantitative and Qualitative Assessment of the Mechanical Properties of the Human Left Ventricle

Computational modeling has contributed to many fields of medicine and has been proven very useful for diagnosing complex diseases. A special case is that of myocardial infarction (MI), a prevalent cardiovascular disease where computational modeling approaches have been useful for detecting abnormalities. The aim of this paper is to propose enhanced computational models for the human left ventricle (LV) to estimate the strain and stress distributions in healthy and diseased subjects. Computational models were developed and evaluated using human LV of 20 patients with MI and 20 healthy controls using cardiac magnetic resonance imaging acquisitions and simulation tools. The finite element technique was employed for LV modeling. Comparative analysis revealed higher global strains in healthy subjects compared to MI patients, particularly global circumferential, longitudinal, and radial strains. The average stress distributions were 67.9 ± 5.01 kPa in healthy models and 78.3 ± 8.21 kPa in infarcted regions. Model-derived strain data indicated an overall average of −0.15 ± 0.06 for healthy models and 0.2 ± 0.04 for infarcted regions. LV strain values were compared with those obtained from two feature-tracking algorithms to validate the proposed models, resulting in very promising findings. The work presented here highlights the importance of computational modeling for quantitative and qualitative analyses of heart disease and the potential of using such models in other organs. When combined with imaging data, the proposed models can have significant implications for improved patient care and treatment strategies.