Efficient estimation of load‐free left ventricular geometry and passive myocardial properties using principal component analysis
| Autor: | Alistair A. Young, Thiranja P. Babarenda Gamage, Zhinuo J. Wang, Vijayaraghavan Rajagopal, Martyn P. Nash, Chris P. Bradley, J Jane Cao, Vicky Y. Wang, Poul M. F. Nielsen |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Hessian matrix
Heart Ventricles 0206 medical engineering Population Biomedical Engineering Diastole 02 engineering and technology 030204 cardiovascular system & hematology 03 medical and health sciences symbols.namesake 0302 clinical medicine Humans Left ventricular geometry Linear combination education Molecular Biology Physics Principal Component Analysis education.field_of_study Estimation theory Myocardium Applied Mathematics Mechanics 020601 biomedical engineering Computational Theory and Mathematics Modeling and Simulation Principal component analysis symbols Identifiability Software |
| Zdroj: | International Journal for Numerical Methods in Biomedical Engineering. 36 |
| ISSN: | 2040-7947 2040-7939 |
| DOI: | 10.1002/cnm.3313 |
| Popis: | Models of cardiac mechanics require a well-defined reference geometry from which deformations and hence myocardial strain and stress can be calculated. In the in vivo beating heart, the load-free (LF) geometry generally cannot be measured directly, since, in many cases, there is no stage at which the lumen pressures and contractile state are all zero. Therefore, there is a need for an efficient method to estimate the LF geometry, which is essential for an accurate mechanical simulation of left ventricular (LV) mechanics, and for estimations of passive and contractile constitutive parameters of the heart muscle. In this paper, we present a novel method for estimating both the LF geometry and the passive stiffness of the myocardium. A linear combination of principal components from a population of diastolic displacements is used to construct the LF geometry. For each estimate of the LF geometry and tissue stiffness, LV inflation is simulated, and the model predictions are compared with surface data at multiple stages during passive diastolic filling. The feasibility of this method was demonstrated using synthetically deformation data that were generated using LV models derived from clinical magnetic resonance image data, and the identifiability of the LF geometry and passive stiffness parameters were analysed using Hessian metrics. Applications of this method to clinical data would improve the accuracy of constitutive parameter estimation and allow a better simulation of LV wall strains and stresses. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Plný text