Material parameter identification for modelling the left ventricle in the healthy state
Master Thesis
2014
Permanent link to this Item
Authors
Supervisors
Journal Title
Link to Journal
Journal ISSN
Volume Title
Publisher
Publisher
University of Cape Town
Department
License
Series
Abstract
An idealized truncated ellipsoidal model, was used to simulate a healthy canine left ventricle. Passive behaviour of the myocardium was modelled using the constitutive model of Usyk. In addition, active behaviour of the myocardium was modelled by the active stress law of Guccione. Furthermore, the load faced by the left ventricle in ejecting blood into the arterial system, was modelled with the three element Windkessel model of Westerhof. The model was calibrated to pressure-volume data, which was adaptedfrom the work of Kerckhoffs. The projected Levenberg-Marquardt algorithm was used to identify material parameters. Identification of the anisotropic constants in the model of Usyk proved to be difficult, with the calibration algorithm often converging to parameter values that produced numerical instability. An idealized truncated ellipsoidal model, was used to simulate a healthy canine left ventricle. Passive behaviour of the myocardium was modelled using the constitutive model of Usyk. In addition, active behaviour of the myocardium was modelled by theactive stress law of Guccione. Furthermore, the load faced by the left ventricle in ejecting blood into the arterial system, was modelled with the three element Windkessel model of Westerhof. The model was calibrated to pressure-volume data, which was adapted from the work of Kerckhoffs. The projected Levenberg-Marquardt algorithm was used to identify material parameters. Identification of the anisotropic constants in the model of Usyk proved to be difficult, with the calibration algorithm often converging to parameter values that produced numerical instability. An idealized truncated ellipsoidal model, was used to simulate a healthy canine left ventricle. Passive behaviour of the myocardium was modelled using the constitutive model of Usyk. In addition, active behaviour of the myocardium was modelled by the active stress law of Guccione. Furthermore, the load faced by the left ventricle in ejecting blood into the arterial system, was modelled with the three element Windkessel model of Westerhof. The model was calibrated to pressure-volume data, which was adaptedfrom the work of Kerckhoffs. The projected Levenberg-Marquardt algorithm was used to identify material parameters. Identification of the anisotropic constants in the model of Usyk proved to be difficult, with the calibration algorithm often converging to parameter values that produced numerical instability. An idealized truncated ellipsoidal model, was used to simulate a healthy canine left ventricle. Passive behaviour of the myocardium was modelled using the constitutive model of Usyk. In addition, active behaviour of the myocardium was modelled by the active stress law of Guccione. Furthermore, the load faced by the left ventricle in ejecting blood into the arterial system, was modelled with the three element Windkessel model of Westerhof. The model was calibrated to pressure-volume data, which was adapted from the work of Kerckhoffs. The projected Levenberg-Marquardt algorithm was used to identify material parameters. Identification of the anisotropic constants in the model of Usyk proved to be difficult, with the calibration algorithm often converging to parameter values that produced numerical instability.
Description
Includes bibliographical references.
Reference:
Essack, M. 2014. Material parameter identification for modelling the left ventricle in the healthy state. University of Cape Town.