Efficient estimation of personalized biventricular mechanical function employing gradient-based optimization.

Area of Science:

• Computational biology
• Cardiovascular mechanics
• Biomedical engineering

Background:

• Accurate measurement of cardiac physiological quantities is challenging.
• Personalized computational models offer a non-invasive estimation method.
• Existing models may lack efficiency and patient-specificity.

Purpose of the Study:

• To develop an efficient framework for patient-specific biventricular heart models.
• To evaluate regional myocardial contractility and myofiber stress.
• To assess the impact of cardiac fiber angles on heart function.

Main Methods:

• Utilized a gradient-based data assimilation method for model personalization.
• Created patient-specific biventricular models from clinical data.
• Performed simulations on a standard laptop, achieving rapid computation times (<2 hours).

Main Results:

• Achieved excellent agreement between simulated and measured cardiac volume and strain data.
• Demonstrated the necessity of steep myocardial fiber angles for accurate motion simulation.
• Quantified significantly lower systolic stresses in the right ventricle compared to the left ventricle.
• Found low variability in estimated parameters across different patients and modeling choices.

Conclusions:

• The developed framework is efficient and accurate for patient-specific cardiac mechanics modeling.
• It enables estimation of crucial clinical parameters like contractility and stress.
• The framework shows promise for clinical diagnostic applications in cardiology.