Professional Documents
Culture Documents
Mathematical Modelling and Electrical Analog Equivalent of The Human Cardiovascular System
Mathematical Modelling and Electrical Analog Equivalent of The Human Cardiovascular System
Mathematical Modelling and Electrical Analog Equivalent of The Human Cardiovascular System
DOI 10.1007/s10558-010-9093-0
ORIGINAL PAPER
Kamran Hassani
Abstract The objective of this study is to develop a Keywords Modelling Cardiovascular system
model of the cardiovascular system capable of simulating Hemodynamics Electrical equivalent
the normal operation of the systemic and pulmonary cir-
culation, starts from aorta, and follows by upper and lower
extremities vessels, finally ends with pulmonary veins. The Introduction
model consists of a closed loop lumped elements with 43
compartments representing the cardiovascular system. The Diseases of the human cardiovascular system are one of the
model parameters have been extracted from the literature. main problems in contemporary health care in the industrial
Using MATLAB software, the mathematical model has countries. These diseases cause the majority of deaths and
been simulated for the cardiovascular system. Each com- also often afflict people in their most productive age. In this
partment includes a Resistor-Inductor-Capacitor (RLC) context different approaches have been used with the aim of
segment. The normal cardiovascular operation is charac- providing better understanding and modelling of the arterial
terised by the pressure–volume curves in different parts of tree and heart mechanism in the human cardiovascular sys-
the system. Model verification is performed by comparing tem. Several mathematical and electrical models dedicated
the simulation results with the clinical observation reported particularly to simulate the operation of cardiovascular
in the literature. The described model is a useful tool in system over the past decade by many scientists such as
studying the physiology of cardiovascular system, and the Beeler and Reuter (1977), Stergiopulos and Young (1992),
related diseases. Also, it could be a great tool to investigate Nebot et al. (1998), Mukkamala and Cotton (2001),
the effects of the pathologies of the cardiovascular system. Migliavacca and Pennati (2001), Heldet and Shim (2002),
Darovic and Stratton (2002), Urquiza and Blanco (2003),
Manoliu (2004), Liang and Liu (2005) and Waite (2006). In
an early effort to model the cardiovascular system, Wester-
hof et al. (1969) developed the Windkessel model to describe
M. Abdolrazaghi, M. Navidbakhsh and K. Hassani contributed the blood flow in the heart. The model included only four
equally to this work. compartments. In later study, he studied the analogue mod-
elling of the human systemic arterial tree. Croston et al.
M. Abdolrazaghi M. Navidbakhsh
(1974) devised a 28-compartment lumped-parameter model.
Mechanical Engineering Department, Iran University of Science
and Technology, Tehran, Iran They modelled arterial, venous, and cardiac compartments
e-mail: Mona.Abdolrazaghi@gmail.com in terms of coupled ordinary differential equations that
M. Navidbakhsh describe the dynamics of the system. In a subsequent
e-mail: m.navid@iust.ac.ir extension of modelling the cardiovascular system, Rupnic
et al. (Rupnic and Runvovc 2002) designed an electronic
K. Hassani (&)
circuit to analyze the operation of cardiovascular system in
Biomedical Engineering Deaprtment, Science and Research
Branch of Islamic Azad University, Tehran, Iran steady-state and transient phenomena. Their model was
e-mail: K.Hasani@srbiau.ac.ir useful in teaching physiology and pathophysiology but the
123
46 Cardiovasc Eng (2010) 10:45–51
arterial system was not detailed. Olufsen and Nadim (2004) can be either linear or time variant. The analogue circuit of
derived a lumped model for the blood flow and pressure in each compartment is actually presented by three-element
the systemic arteries using the equations of fluid dynamics. Windkessel model in Fig. 2a, b. The description of viscous
They started with the one-dimensional Navier–Stokes resistance (R1) (R2) defined according to the following rela-
equations describing time-dependent flow and pressure in a tions (Olufsen and Nadim 2004):
rigid vessel and derived first and second–order lumped
models using Laplace transform. The resulting lumped kq 8lL 8lL
R1 ¼ 1 ; R2 ¼ ð1Þ
model can be represented by electrical circuits. Hassani et al. kp p:r 4 p:r 4
(2006) presented a new model for normal operation of car-
The values of blood density (q) and viscosity (l) as well as
diovascular system and implemented a 42-compartment
length of artery (L), radius (r), thickness (h) and Young
steady state model of the system connected to models of left
modulus (E) have been extracted from literature (Wang and
and right ventricles. Each compartment included a resistor,
Parker 2004). The inertance (L1), due to blood inertia, was
capacitor and inductor. The model described the arterial
described as follows (Olufsen and Nadim 2004):
system in details and was used to simulate the cardiovascular
system using an equivalent electronic system. Using OR- 8qL
L1 ¼ ðkq kp Þ ð2Þ
CAD software, we designed a circuit for the operation of the p r2
system in normal condition. Despite the fact that we tried as The elastic behaviour of the artery was described by the
much as possible to avoid a passive arterial model, we compliance (C), determined under the following relation
acknowledge that our previous model is not active. In an (Rideout 1991):
attempt to overcome the limitation of passive cardiovascular
3p r 3 L
models, we have taken a different approach to the modelling C¼ ð3Þ
2Eh
of the system and tried to improve our previous model by
investigating the stiffness of the ventricles and presenting The input flow is represented by blood flow (q) through the
precise pressure–volume graphs for different parts of the left ventricle which was described by (Olufsen and Nadim
cardiovascular system. Also, we have extracted a new 2004):
pressure-flow mathematical model describing the flow in
1 1 dq 1 L1 dp
vessels based on a vessel diameter. The system has been L1 þ þq¼ þp ð4Þ
simulated by MATLAB software, and all the results R1 R2 dt R2 R1 dt
including the pressure–time and volume-time graphs of the Note that the pressure (p) actually represents the pressure
system segments have been compared with the relevant difference between the inlet and outlet of the artery. The
clinical data. Moreover, these results were verified by values of time-scale kp and kq depend on the radius of the
physiological and experimental data. artery (r). If the radius of the artery is less than 0.2 cm, the
compartment has only one resistor and there are two resistors
in the compartment for the arteries with the radius of more
Methods than 0.2 cm. Furthermore, the value of time-scale is
determined according to the following relations (Olufsen
We have extended our previous model to facilitate represen- and Nadim 2004):If radius of the artery is less than 0.2 cm
tation of cardiovascular system operation in normal condition then:
and presented a mathematical model with active elements.
kq ¼ 0:1729; kp ¼ 0:0075
The model is mathematically formulated in terms of an elec-
tric circuit. Figure 1 describes the block diagram of the car- If radius of the artery is more than 0.2 cm then:
diovascular system staring from ascending aorta and ends with
kq ¼ 0:2057; kp ¼ 0:0392
pulmonary veins. This model can be divided into two main
parts: the heart (left and right), including atriums and ventri- The values of the hemodynamic parameters for ventricles,
cles, and the arterial system. To facilitate the understanding of pulmonary artery, veins, arterioles, capillaries, and atriums
the system, we have depicted the anatomic structures into the have been extracted from the literatures (Rideout 1991).
diagram. The peripheral circulation is divided into upper body The entire electrical equivalent of the model is shown in
(including Carotid and Subclavian arteries), Aorta artery, Fig. 3. The electronic parameters are correlated to their
Gastric, Hepatic, Splenetic, Renal, Mesenteric branches and mechanical parameters as follows: voltage (volt) is analo-
lower extremely sections. Arterioles and Capillaries and veins gous to pressure (mmHg), capacitance (lF) to compliance
are separately identified. The model consists of 43 compart- (ml/Pa), resistance (kX) to resistance (1 Pa.s/ml), and
ments, each of the compartments is represented by one or two inductance (1lH) to inertance (1 Pa.s2/ml). The heart rate is
linear resistances, one linear inductor and one capacitor that considered 75 beats per minute and the length of cardiac
123
Cardiovasc Eng (2010) 10:45–51 47
cycle is 0.8 s. The pumping action of the heart is realized by 2p
varying the right and left ventricular capacitors according to SRV ¼ 1=CRV ¼ SRS sin þ SRD UðtÞ ð6Þ
T
a step-function U(t) with unit amplitude. The function has a
pulse width of 35.5% indicating the ratio of the systolic SRS, SRD represent the right ventricle stiffness in systole
duration to the cardiac cycle (0.3 s/0.8 s). The defined unit- and diastole. CRV denotes the compliance of right ventricle.
step function is a time dependent function for the ventricles We considered (Rideout 1991) the atrium pressure
to show that the heart signals are pulsated. We have calcu- (PAT) = 4.5 mmHg, left ventricle pressure in systole
lated the left ventricle stiffness in our model using the fol- (PES = 90 mmHg), maximum left ventricle volume
lowing relation: (QLVmax ¼ 120 ml) and minimum left ventricle volume
(QLVmin ¼ 48 ml); to calculate the left ventricle systolic and
1 2p
SLV ¼ =CLV ¼ SLS sin þ SLD UðtÞ ð5Þ diastolic stiffness as follows:
T
SLD ¼ PAT 1332=QLVmax ¼ 4:5 1332=120
SLV denotes the total left ventricle stiffness, SLS and SLD ¼ 49:5 gm=cm4 s2 ð7Þ
represent the left ventricle stiffness in systole and diastole. CLV
expresses the compliance of left ventricle. In a similar method, SLS ¼ PES 1332=QLVmin ¼ 90 1332=48
we calculated the total right ventricle stiffness (SRV): ¼ 2497:5 gm=cm4 s2 ð8Þ
123
48 Cardiovasc Eng (2010) 10:45–51
Fig. 3 The lumped model of the cardiovascular system. The anatomical name of each element has been included
123
Cardiovasc Eng (2010) 10:45–51 49
Fig. 7 The pressure graph of right atrium (PRA) and systemic vein II Fig. 9 The pressure graph of left atrium (PLA) and pulmonary vein
(PSYSVEIN2) II (PPULVEIN2)
123
50 Cardiovasc Eng (2010) 10:45–51
123
Cardiovasc Eng (2010) 10:45–51 51
describing the system based on the vessel diameters, and Manoliu V (2004) Consideration about the lumped parameter
simulating the mathematical equations with active electri- windkessel model applicativity in the cardiovascular system
structure, National symposium of theoretical electrical engineer-
cal elements. We have outlined the use of the model as a ing, 22–23, Burcharest, UPB, pp 428–437.
tool to better understand the physiology of the system and Migliavacca F, Pennati G. Modeling of the Norwood circulation
study the related diseases. effects of shunt size, vascular resistance and heart rate. Am J
Physiol Heart Circ Physiol, Am J Physiol Heart Circ Physiol.
Acknowledgments The authors would like to acknowledge all the 2001;280:457–70.
help and support of their colleagues at Iran University of science and Mukkamala R, Cotton RJ. A forward model based validation of
technology and Chagalesh Consulting Engineers Company. cardiovascular system identification. Am J Physiol Heart Circ
Physiol, Am J Physiol Heart Circ Physiol. 2001;281:2714–30.
Nebot A, Cellier FE, Vallverdu M. Mixed quantitative/qualitive
modeling of the cardiovascular system and simulation of it. J
References Comput Methods Programs Biomed. 1998;55:127–55.
Olufsen MS, Nadim A. On deriving lumped models for blood flow
Beeler GW, Reuter H. Reconstruction of the action potential of the and pressure in the systemic arteries. J Math Biosci Eng.
ventricular myocardial fibers. J Physiol. 1977;268:177–210. 2004;1(1):61.
Croston RC, Fitzjerrell DG. Cardiovascular model for the simulation Rideout VC. Mathematical and computer modeling of physiological
of exercise, lower body negative pressure, and tilt table system. New Jersey: Printice-Hall Inc; 1991. p. 27–66.
experiments. In: Proceedings 5th Annual Pittsburgh conference Rupnic M, Runvovc F. Simulation of steady state and transient
modeling simulation, 1974, pp 471–476. phenomena by using the equivalent electronic circuit. J Comput
Darovic OG, Stratton KL. Hemodynamic monitoring, invasive and Methods Programs Biomed. 2002;67:1–12.
non invasive clinical application. 3rd ed. Saunders: Philadelphia; Stergiopulos N, Young D. F Computer simulation of arterial flow with
2002. p. 306–39. applications to arterial and aortic stenosis. J Biomed.
Guyton AC. Overview of circulation; medical physics of pressure, 1992;25:1477–88.
flow, and resistance. In: Hall JE, editor. Textbook of medical Urquiza S, Blanco P. Coupling multidimensional compliant model for
physiology. 9th ed. Pennsylvania: WB Saunders; 1996. p. 107– carotid artery blood flow. J Mecánica Computacional.
19. 2003;XXII:232–43.
Hassani K, Navidbakhsh M, Rostami M (2006) Simulation of the Waite L. Biofluid mechanics in cardiovascular system. Mc Graw-Hill,
cardiovascular system using equivalent electronic system, J 2006.
biomedical Papers of the medical faculty of the University Wang JJ, Parker KH. Wave propagation in a model of arterial
Palacky, Olomouc, 150(1): 105–112. circulation. J Biomech. 2004;37:457.
Heldet T, Shim E. Modeling of cardiovascular response to orthostatic Westerhof N, Bosman F, DeVries CJ, Noordergraaf A. Analogue
stress. J Appl Physiol. 2002;92:1239–54. studies of the human systemic arterial tree. J Biomech.
Liang F, Liu H. A closed-loop lumped parameter computational 1969;2:121–43.
model for human cardiovascular system. JSME Int J.
2005;48(4):484. Series C.
123