Establishment of a 4D Computational Model for Wall Stress Calculation in Healthy and Diseased Hearts, Using a Rat Infarction Model

Objectives: Heart failure (HF) in pediatric patients is often a progressive disease that is difficult to diagnose at the beginning. Myocardial wall stress increases early at the beginning of HF and is influenced by different parameters such as wall thickness and intraventricular pressure development.

One of the most effective approaches for investigating insights into the heart pathology is to use computational models. To adapt a wall stress model active and passive mechanical properties of cardiac muscle tissue of healthy and diseased hearts as well as the changing of the heart morphology (4D) during one heart cycle are measured. This study demonstrates the establishment of a 4D computational model, which is able to simulate myocardial wall stress in healthy and infarcted hearts, using 3D MRI and intraventricular pressure information.

Methods: Sprague-Dawley rats were divided into a control (n = 8) and a HF group (n = 8). HF was induced by permanent ligation of left anterior descending coronary artery. Two weeks after cardiac infarct 4D high-resolution CMRT (7T) and intraventricular pressure measurement were performed. Hearts were extracted afterward and passive tissue mechanic parameters were measured in different cardiac regions, using an automated biomechatronic system (MyoRobot).

Mathematical algorithms were implemented in a finite element-based computational model. Simulation was started by inserting the CMRT-measured 3D heart end diastolic (ED) volume.

A comparison was performed of simulation-calculated heart morphology and heart morphology registered by CMR in two different heart cycles (ED/end systolic [ES]) using the dice similarity coefficient (DSC). Local wall stress was shown in 4D and the average was output for each time point of the heart cycle.

Results: The simulation of healthy rats shows a very close similarity to real measured 3D morphology in both heart cycles (DSC-ED 0.89 ± 0.04, DSC-ES 0.93 ± 0.05). Infarcted hearts also show a strong similarity, but the variability was much greater in those rats (DSC-ED 0.86 ± 0.09, DSC-ES 0.90 ± 0.12). Computed average wall stress between healthy and infarcted hearts shows a significant difference in ED and ES (p < 0.01)

Conclusion: 4D computer simulation including wall stress information based on real measured 3D cardiac morphology in healthy and infarcted hearts is a promising method to further investigate the pathology of beginning HF.