Robust and efficient fixed-point algorithm for the inverse elastostatic problem to identify myocardial passive...

Area of Science:

• Computational Biomechanics
• Medical Imaging Analysis
• Cardiac Physiology

Background:

• Personalized computational heart models are crucial for clinical translation.
• Existing methods for characterizing myocardial passive behavior have limitations, especially with high-resolution geometries.

Purpose of the Study:

• To present a novel, automated methodology for identifying *in vivo* passive cardiac biomechanics properties.
• To enable the creation of personalized cardiac models using clinical data.

Main Methods:

• Developed a highly-efficient algorithm to fit material parameters against patient-specific end-diastolic pressure-volume relations (EDPVR).
• Implemented a novel line search strategy for generating an unloaded reference configuration, enhancing convergence and robustness.
• The method requires only clinical image data or meshes and one EDPVR data point.

Main Results:

• The algorithm successfully automates the identification of passive cardiac biomechanics properties.
• The generated unloaded reference configuration is robust and converges efficiently.
• Sensitivity analysis confirmed the algorithm's robustness to initial input parameters.

Conclusions:

• The proposed method offers a robust and efficient approach for personalizing cardiac biomechanical models.
• It can be readily integrated with existing finite element software packages.
• This facilitates more accurate diagnosis and therapy planning for cardiac conditions.