DocumentCode
3562179
Title
Quantitative insights into the closed loop cardiovascular system using an electrical lumped element physiological model
Author
Tsanas, Athanasios ; Clifford, Gari D. ; Vartela, Vassiliki ; Sfirakis, Petros
Author_Institution
Univ. of Oxford, Oxford, UK
fYear
2014
Firstpage
509
Lastpage
512
Abstract
This study extends a comprehensive closed-loop physiological model of the circulatory system based on lumped electrical analogue components. The model associates subject specific factors such as age, gender, body surface area, fitness and smoking habits, with cardiovascular parameters including vascular blood pressures, blood volumes, and cardiac output, providing physiological insight via the interpretation of the model´s parameters. The model parameters were determined using multi-objective constrained optimization in a typical inverse problem setup, and were subsequently associated with the subject specific factors using least squares polynomial relationships with ℓ1-ℓ2-norm regularization. We validated the performance of the model using data from 289 subjects, replicating arterial blood pressure accurately (about 4.8% relative deviation from the measured values), whilst also providing physiologically realistic estimates of vascular blood pressures, blood volumes, and cardiac output.
Keywords
cardiovascular system; closed loop systems; equivalent circuits; inverse problems; least squares approximations; lumped parameter networks; optimisation; physiological models; polynomial approximation; arterial blood pressure; blood volume; cardiac output; cardiovascular parameter; circulatory system; closed loop cardiovascular system; electrical lumped element physiological model; inverse problem setup; least squares polynomial relationship; lumped electrical analogue component; multiobjective constrained optimization; vascular blood pressure; Abstracts; Computational efficiency; Immune system; Physiology; Resistance; Standards;
fLanguage
English
Publisher
ieee
Conference_Titel
Computing in Cardiology Conference (CinC), 2014
ISSN
2325-8861
Print_ISBN
978-1-4799-4346-3
Type
conf
Filename
7043091
Link To Document