RT Journal Article
T1 Efficient cardiovascular parameters estimation for fluid-structure simulations using gappy proper orthogonal decomposition
A1 Deus, J.
A1 Martín Ortega, Elena Beatriz
K1 2204.04 Mecánica de Fluidos
K1 1299 Otras Especialidades Matemáticas
AB As full-scale detailed hemodynamic simulations of the entire vasculature are not feasible, numerical analysis should be focused on specific regions of the cardiovascular system, which requires the identification of lumped parameters to represent the patient behavior outside the simulated computational domain. We present a novel technique for estimating cardiovascular model parameters using gappy Proper Orthogonal Decomposition (g-POD). A POD basis is constructed with FSI simulations for different values of the lumped model parameters, and a linear operator is applied to retain information that can be compared to the available patient measurements. Then, the POD coefficients of the reconstructed solution are computed either by projecting patient measurements or by solving a minimization problem with constraints. The POD reconstruction is then used to estimate the model parameters. In the first test case, the parameter values of a 3-element Windkessel model are approximated using artificial patient measurements, obtaining a relative error of less than 4.2%. In the second case, 4 sets of 3-element Windkessel are approximated in a patient’s aorta geometry, resulting in an error of less than 8% for the flow and less than 5% for the pressure. The method shows accurate results even with noisy patient data. It automatically calculates the delay between measurements and simulations and has flexibility in the types of patient measurements that can handle (at specific points, spatial or time averaged). The method is easy to implement and can be used in simulations performed in general-purpose FSI software.
PB Annals of Biomedical Engineering
SN 00906964
YR 2024
FD 2024-07-05
LK http://hdl.handle.net/11093/7176
UL http://hdl.handle.net/11093/7176
LA eng
NO Annals of Biomedical Engineering, Pending(undefined): 1-16 (2024)
NO Universidade de Vigo/CISUG
DS Investigo
RD 11-dic-2024