Download Review on CFD simulation in heart with dilated cardiomyopathy and

Computers in Biology and Medicine 43 (2013) 377–385
Contents lists available at SciVerse ScienceDirect
Computers in Biology and Medicine
journal homepage: www.elsevier.com/locate/cbm
Review on CFD simulation in heart with dilated cardiomyopathy
and myocardial infarction
Bee Ting Chan a,n, Einly Lim a,1, Kok Han Chee b,2, Noor Azuan Abu Osman a,3
a
b
Department of Biomedical Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia
Department of Medicine, Faculty of Medicine Building, University of Malaya, 50603 Kuala Lumpur, Malaysia
a r t i c l e i n f o
abstract
Article history:
Received 12 April 2012
Accepted 20 January 2013
The heart is a sophisticated functional organ that plays a crucial role in the blood circulatory system.
Hemodynamics within the heart chamber can be indicative of exert cardiac health. Due to the
limitations of current cardiac imaging modalities, computational fluid dynamics (CFD) have been
widely used for the purposes of cardiac function assessment and heart disease diagnosis, as they
provide detailed insights into the cardiac flow field. An understanding of ventricular hemodynamics
and pathological severities can be gained through studies that employ the CFD method. In this research
the hemodynamics of two common myocardial diseases, dilated cardiomyopathy (DCM) and myocardial infarction (MI) were investigated, during both the filling phase and the whole cardiac cycle,
through a prescribed geometry and fluid structure interaction (FSI) approach. The results of the
research indicated that early stage disease identification and the improvement of cardiac assisting
devices and therapeutic procedures can be facilitated through the use of the CFD method.
& 2013 Elsevier Ltd. All rights reserved.
Keywords:
Computational fluid dynamics
Dilated cardiomyopathy
Myocardial infarction
Contents
1.
2.
3.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
Clinical measurements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
CFD approaches to simulate Blood flow in DCM and MI conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
3.1.
Dilated cardiomyopathy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
3.2.
Myocardial infarction (MI). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
4. Future direction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Conflict of interest statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Acknowledgment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
1. Introduction
Heart failure is characterized by the inability of the heart to supply
adequate blood flow and therefore deliver oxygen to tissues and
organs in the body. It is usually induced by cardiovascular diseases
n
Corresponding author. Tel.: þ60 1 6752 3764; fax: þ60 3 7967 7661.
E-mail addresses: [email protected] (B.T. Chan),
[email protected] (E. Lim), [email protected] (K.H. Chee),
[email protected] (N.A. Abu Osman).
1
Tel.: þ60 1 2212 3632; fax: þ60 3 7967 7661.
2
Tel.: þ60 1 7695 2750; fax: þ 60 3 7953 5627.
3
Tel./fax:þ 60 3 7967 7661.
0010-4825/$ - see front matter & 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.compbiomed.2013.01.013
such as coronary artery disease, arrhythmias, congenital heart disease,
heart valve disease, congestive heart disease, rheumatic heart disease,
stroke and high blood pressure. Cardiovascular disease is the most
commonly reported cause of mortality and it contributes to approximately 30% of deaths worldwide [1]. When an individual suffers from
cardiovascular disease a number of compensatory mechanisms take
place in order to maintain cardiac output and these include geometrical modifications of the heart. The heart is stretched to hold more
blood in the diastole so that it is able to generate stronger force of
contraction during systole, following Frank–Starling’s law [2–4].
Furthermore, to overcome the high afterload pressure that is normally experienced by patients that are suffering from heart disease,
yet with reasonable wall stress, the muscle wall thickens. Over a long
378
B.T. Chan et al. / Computers in Biology and Medicine 43 (2013) 377–385
period the gradual declination of cardiac performance is no longer
compensable. At this stage, clinical symptoms such as fatigue,
dizziness, diminished exercise capacity, shortness of breath and
edema are observed.
Heart failure patients may experience systolic dysfunction,
diastolic dysfunction, or both. Systolic dysfunction refers to the
abnormal performance of the heart caused by insufficient contraction, while diastolic dysfunction refers to abnormalities that
are caused by the insufficient relaxation of the heart. Systolic
dysfunction involves a progressive condition which leads to
cardiac remodeling, which is characterized by dilatation, changes
in sphericity, wall thinning, decreased cardiac reserve, impaired
exercise tolerance, increased wall stress [5] and thus increased
myocardial oxygen demand [6]. DCM is one of the most common
cardiac diseases which exhibits systolic dysfunction. Diastolic
dysfunction, on the other hand, is characterized by slow relaxation of the heart and abnormal filling patterns, most often caused
by increased stiffness of the cardiac muscle [7–8]. Diastolic heart
failure is shown in hypertrophy cardiomyopathy where the heart
develops thicker and stiffer heart muscles that show signs of
impaired relaxation.
Numerous methods have been used to diagnose and differentiate various types of heart failure conditions in order to devise
the best treatment strategies for the patients involved. These
involve examining the heart’s morphology [9–11], electrical
activity [12–13], mechanics [11,14–15] and hemodynamics
[16–17]. Invasive diagnostic methods, such as blood test and
coronary catheterization procedures, are routinely used in a
clinical setting. With the advancement of medical technologies,
noninvasive imaging modalities, such as chest X-rays, electrocardiograms (ECG), computed tomography (CT) and magnetic
resonance imaging (MRI) are gradually becoming more popular.
Among these, MRI and echocardiography are the most commonly
used diagnostic tools that are used to assess cardiac function
through geometric and flow measurements, such as left ventricle
(LV) volume, wall mass, stroke volume, ejection fraction (EF), wall
motion and wall thickness. However, differentiated velocity
vector and pressure field, as well as the local hemodynamics
indices such as mass transport, wall shear and boundary flow
layer [18], which are important parameters for early diagnosis of
heart failure, are not able to quantitatively evaluate the effects of
individual parameters to the disease conditions through imaging
modalities.
The CFD method involves the study of cardiovascular blood
flow patterns and it has emerged as a reliable tool that can be
used to enhance our understanding of the pathophysiology and
progression of heart disease by providing the means through
which reproducible numerical experiments can be produced
under identical conditions. Global and regional hemodynamics
variables, such as intraventricular blood flow dynamics, ventricular wall motion, spatial and temporal distributions of pressure
and myocardial strains and stresses, can be obtained through the
simulations. Information about these parameters provides opportunities for the early diagnosis of certain heart diseases while
sensitivity analysis performed through the CFD method is able to
demonstrate the correlation between individual parameter to the
disease condition. Early CFD techniques for hemodynamics simulations were mostly carried out on simplified geometries [19–21].
With the development of cardiac imaging techniques, patientspecific morphology and flow have been progressively used
[22–25], and this provides valuable clinical information. In the
recent decade, FSI models that take into account the interaction
between the blood and the cardiac wall have been developed
[26–30].
The present paper provides a comprehensive overview of the
existing diagnostic methods, including CFD simulations, in terms
of their ability to identify the presence of the two most common
myocardial diseases, i.e. DCM and MI during filling phase. Currently available and commonly used diagnostic tools are presented, with an emphasis on the various information or
parameters they provide, as well as their limitations on global
variables. This is followed by a review of the existing CFD models
of the diseases, focusing on their methods, findings (the global
and regional hemodynamic variables), as well as validations of
the results.
2. Clinical measurements
As a result of heart failure a series of examinations are carried
out for the purpose of inspecting cardiac function. Cardiac
morphology is noninvasively measured through CT, MRI and
echocardiography. The parameters, such as wall thickness, end
diastolic dimension (EDD) and end systolic dimension (ESD), are
acquired in linear or volumetric measurements. In cardiac
mechanics assessment, the LVEF is the gold standard for cardiac
systolic function [31]. This global index represents ventricular
contraction strength, which is easily obtained through MRI and
2D echocardiography. Myocardial contractility is usually
inspected through wall motion, which is detected in tissue
Doppler imaging. The diastolic function is currently indicated by
deceleration time and E/A ratio [32], which are normally attained
during the LV filling phase through pulsed wave Doppler echocardiography. To some extent, M-mode Doppler and MRI flux
measurements provide hemodynamics information such as flow
velocity propagation.
The hallmark of DCM is ventricular dilatation with myocardial
contractile dysfunction [33], where reducing wall thickness and
ventricular chamber enlarge around 4 cm yields spherical ventricular shape. Commonly under measurement of MRI [34] and
echocardiography [35–37], it shows increased EDV (greater than
112%) and ESV with thinner wall in anatomical data, while
functionally with low EF (less than 45%) and decreased wall
motion with fractional shortening less than 25%. Mitral regurgitation flow is considered to be a hemodynamic disorder [37].
MI is indicated by a bulge found at the affected area, thus
resulting in an irregular ventricular shape under MRI [38–40] and
echocardiograph [41–42]. The bulge area has a wall thickness 30%
thinner than the adjacent segment [43] and this is usually less
than 5.5–6 mm. Due to regional contractility loss, MI shows
slightly lower EF compared to global effect on DCM. However,
MI may develop into DCM due to the over stretching of the
muscle at normal region to compensate infarcted region mechanism [44]. Meanwhile hemostasis region is observed as an identification of abnormal blood flow pattern [45–46].
Although global LV function depends on both systolic and
diastolic functions, the findings from imaging modalities conclude
systolic dysfunction in DCM and MI through EF reduction [47–48],
while the gauge of diastolic dysfunction is not yet well-found.
Although E/A ratio has been widely used to assess diastolic
dysfunction, recent studies [49–51] question its efficiency in
certain disease conditions. The pseudonormal condition shows
normal E/A ratio due to the effects of impaired relaxation and
restriction, which compensate each other. It is therefore highly
likely that this can lead to an incorrect diagnosis. Hereby
hemodynamics and combination imaging techniques [36,52] are
required to reveal the abnormalities.
Yet, the evaluation of cardiac function has traditionally been
limited to global geometric measurements which restrict the
information based on the regional assessments due to temporal
and spatial resolution constraint. Regional measurement and
hemodynamics play crucial roles in disease initiation and
B.T. Chan et al. / Computers in Biology and Medicine 43 (2013) 377–385
progression. In advanced MR function, measurements pertaining
to the regional myocardial functions are available through MR
tagging [53], however it is a time consuming process.
3. CFD approaches to simulate Blood flow in DCM
and MI conditions
The CFD approach provides approximate solution of velocity
and pressure fields through Navier–Stokes equation [54]. In terms
of its strength over current imaging diagnostic tools, CFD further
provides hemodynamics factors such as intraventricular flow
dynamics, wall shear, mass transport and stagnation region
[55]. The velocity and vorticity distribution have a complicated
relationship in thrombus formation [56], which is important
during the identification of cardiac disease. Hereby various
physical flow conditions can be predicted, while inducing the
early recognition of disease stage helps to reduce patient’s risk. As
a result, CFD simulation has been extensively applied as a means
of investigating ventricle pathological flow [57–58] and evaluating surgical treatment and cardiac assist devices. Existing studies
that examine CFD as a means of examining the common myocardial diseases of DCM and MI are discussed in this review study.
3.1. Dilated cardiomyopathy
The blood flow distribution within dilated LV was studied during
the filling phase [59] and the whole cardiac cycle [60]. LV was
represented in a simple axisymmetrical geometry (spheroid). Hemodynamics at different stages of DCM were analyzed and compared
with that of a healthy LV. During the filling phase study [59], LV was
characterized with dimension D representing diameter and H
379
representing height of the left ventricle. The different stages of
DCM were modeled according to end diastole linear relation
D ¼ 0:04a þ 0:05 ½m;
H ¼ 0:02aþ 0:09 ½m
formulated through clinical data. The value of a varies from 0
(healthy) to 1 (severe dilatation). The author simulated a healthy
heart as well as three different stages of DCM condition to study the
relationship between LV dilatation and flow distribution. The blood
flows into the left ventricle through the mitral orifice interact with
the left ventricular wall. The pulsatile flow rate obtained from
clinical data is applied as the inlet boundary condition. The wall
motion is prescribed with change of diameter and height of LV
according to the pulsatile flow rate.
In healthy LV, two vortex rings are formed during the filling
phase, where the primary vortex ring grows in size and propagates toward the apex during early of diastole and a smaller
secondary vortex ring is then formed at basal region at late of
diastole. High propulsion energy is obtained through great velocity and vorticity strength. In dilated LV, only one vortex ring is
observed and it remains attached to the valvular edge throughout
the filling phase (Fig. 1). The attachment is induced by wake
formation, which slows down vortex impulse and propagation
toward the apex. When the volume of LV dilatation increases, the
size of the vortex ring becomes larger with lower propagation
velocity. The stagnant flow is observed in the lower apical region,
which prompts the condition of thrombosis.
The existence of DCM leads to continuous deterioration of the
cardiac function. The left ventricular assisted device (LVAD) was
invented as a means of improving poor cardiac function by
regulating flow within the LV. The role that LVAD plays in supporting DCM heart was investigated throughout the cardiac cycle [60]. A
simplified ellipsoid geometry was connected to cylindrical tubes at
Fig. 1. Flow distribution in healthy LV (left) with A¼ 0 at t ¼3/64 (top), t¼ 6/64 (middle) and t ¼21/64 (bottom); dilated LV(right) with A ¼ 0.25, t¼ 15/64 (top), A ¼ 0.5,
t ¼20/64 (middle) and A ¼ 0.7, t ¼22/64 (bottom) [59].
380
B.T. Chan et al. / Computers in Biology and Medicine 43 (2013) 377–385
Fig. 2. PV loops for healthy and DCM ventricles [60].
both the top (inlet) and bottom (outlet). A lumped parameter model
consists of systemic and pulmonary circulation that is made of
ventricular chamber, valves, venous and arteries in electrical circuit.
A LVAD was modeled by connecting a flow pump between the LV
and the aorta. The volume change in the LV was derived from a
circulatory model and used to obtain fluid or wall velocity.
Aside from the healthy LV, end stage DCM and assisted DCM
with various LVAD flow rates were simulated for the purpose of
comparison. The end stage of DCM has large ventricular volume
and lower ejection fraction compared to healthy LV. With increasing flow rate in LVAD, it was found that ventricular volume
decreases and ejection fraction increases, which shifts the
pressure–volume (PV) loop to the left (Fig. 2).
Instead of two vortex rings that were observed in the first
study [59], another third vortex ring was then observed in the
Fig. 3. Flow fields in healthy LV (top), DCM (LVAD flow¼ 0 L/min) (middle) and DCM (LVAD flow¼ 6 L/min) (bottom): (a) maximum filling rate; (b) end of filling phase;
(c) maximum ejection rate; and (d) end of ejection phase [60].
B.T. Chan et al. / Computers in Biology and Medicine 43 (2013) 377–385
healthy LV during the end of filling. This vortex rings developed
and rotated in opposite directions as a result of the interaction
between the first and second vortex rings and the wall of the LV.
The vortex ring then disappeared during the ejection period
(Fig. 3: top).
However, with end-stage DCM, the LV exhibited lower mitral
flow and smaller vortex rings were observed at the beginning of
filling. The vortex ring stayed near the apex during ejection and a
small part of the next filling phase (Fig. 3: middle). The vortex
ring size deviated from the first study because the models were
built with different DCM severity, where the end-stage DCM
model built was too weak and this entailed that the incoming
jet was no longer able to induce a strong vortex, hence dilatation
would not produce the larger vortex.
With the assistance of the LVAD in the DCM heart, the flow
field was corrected to replicate that of a normal heart (Fig. 3:
bottom). The increase in the LVAD flow rate decreased the volume
of the LV and, as such, vortex formation was more effective. The
vortex strength, area and vorticity were also improved.
Fig. 4. Distribution of normal, infarcted and border zone region [66].
381
3.2. Myocardial infarction (MI)
FSI analysis was used to investigate the influence of MI on LV
hemodynamics. A LV was built that had a composition of 40%
infarction region at apex, 10% at the border zone region and 50%
at normal region at the base (Fig. 4).
The properties of infarcted tissue change with time, where the
stiffness initially reduces and then progressively increases after
days or weeks of infarction. The contractility of infarcted and
border zone region is 30% and 70% of the normal region respectively. The same preload and afterload were applied for all the
conditions as the objective of the study was to investigate
differences in mechanical behavior between a normal and a
diseased heart, excluding compensatory mechanism.
In the LV with MI, the myocardial wall at the normal region
becomes thick while it is activated, the infracted region is
inactivated and stretched to become thinner and bulged (ventricular aneurysm), as shown in Fig. 5. Flow velocities are lower in
the ventricular aneurysm, 10 cm/s during systole and 5.16 cm/s
during diastole compared to velocity of systole 15.6 cm/s and
diastole 9.1 cm/s at the normal region. This suggests that there is
a causal relationship between aneurysm formation and ventricular thrombus via blood stagnation.
The flow abnormality, intraventricular flow and pressure
gradient were analyzed throughout the cardiac cycle in normal
and acute infarction left ventricle (Fig. 6). During systole (0.2 s),
the whole normal LV filled with ejection flow directing to
aortic valve and a monotonous pressure gradient was observed
from the apex to the aortic outflow. However, the ejection flow
in the infarcted left ventricle existed only in the upper half, even
the flow directing toward apex was observed. Higher pressure in
the upper portion of LV was detected because myocardial tissue
in the normal region generates contraction force. During diastole
(0.8 s), large vortex was observed in both the infarcted and
healthy LV. However in the LV with MI, the vortex shape was
distorted and presented a more rectangular shape where its
center shifts toward the apex.
Meanwhile, the energetic efficiency for the LV models was
investigated. The energy efficiency for the healthy LV, subacute
and acute MI was 82.1%, 71.2% and 68.3% respectively. It can be
concluded that the LV with MI is less efficient than that of a
healthy LV.
MI is often closely related to wall motion abnormalities due to
impaired contractility of injured muscle. The decreased wall
Fig. 5. Intraventricular flow pattern and wall displacement [66].
382
B.T. Chan et al. / Computers in Biology and Medicine 43 (2013) 377–385
Fig. 6. Intraventricular flow pattern and pressure distribution during systole (0.2 s) and diastole(0.8 s) in normal and infarcted LV [66].
motion can be classified as hypokinetic (reduced wall motion),
akinetic (absent wall motion) and dyskinetic (bulge). Thus the
relationship between wall motion abnormalities and intraventricular fluid dynamics is of more interest to researchers [61]. LV
was modeled as a truncated prolate spheroid to provide basic
insight into ventricular flow. The ventricular wall motion is
formulated by reducing longitudinal and circumferential strain
from healthy to infarcted LV.
At instant of time, healthy strain at every point was multiplied
by reduction coefficient to reduce contraction,
" #
yyc 4 ZZc 4
cðZ, yÞ ¼ 1A exp 2ys
2Zs
where infarction region has center at yc ¼ 5p=6 and Zc ¼ p=8,
and extension ys ¼ p=3 and Zs ¼ p=10; factor A represents infarction entity, where A¼0 is a healthy LV without infarction, while
regional akinesia is represented by A¼1 and A¼ 2 denotes dyskinetic
condition.
The sets of results were analyzed in terms of vorticity and
velocity. During diastole, vortex rings were observed with primary vortex deeply penetrating into the LV. During early diastole,
compact vortex rings with two counter-rotating vortices were
observed. The anterior vortex occupied a large part of the LV
center and pushed the posterior vortex (Fig. 7: top (a)) creep
along posterior wall that interacted with the viscous boundary
layer, which partly dissipated within the boundary layer. At late
of diastole, a secondary vortex was shaped in the basal region.
During systole, vortices interacted and produced complex 3D field
(Fig. 7: top (d)) which then dissoluted and ejected.
In the LV with akinetic wall motion, the vortex penetrated less
deeply into the LV and remained in the basal region during
diastole (as shown in Fig. 7: middle (a)). It had a lower level of
vorticity and strength. During systole, the flow had low velocity
and a more regular vortex was exposed (Fig. 7: middle (c)) due
to weaker self-induced dynamics and dissipation. The ejection
fraction reduced from 55% acquired in the healthy LV to 45%.
The dyskinetic wall motion had a lower level of vorticity with
increased shortening of mitral jet (Fig. 7: bottom). The small
vortex is formed in the upper basal region. During systole (Fig. 7:
bottom (c)), flow redirected to aortic outflow from the central and
upper regions of the LV, while null velocity was spotted close to
the apex area. This yielded a much lower ejection fraction of 32%.
This study ascertained that flow was stagnant at the apex and
close to the wall segment with reduced mobility due to insufficient vortex strength to wash out the fluid. Hence blood clot
formation or thrombosis tends to take place in the infarcted LV.
Surgery is one option for the treatment of myocardial infarction. An investigation into the effect of surgical ventricle reconstruction (SVR) on patient-specific ventricle was carried out [62].
The cardiac MR images of the MI patient were transformed into
grid images and applied in Karlsruhe Heart Model (KaHMo) [63]
B.T. Chan et al. / Computers in Biology and Medicine 43 (2013) 377–385
383
Fig. 7. Flow in healthy LV (top), LV with moderate regional akinetic (middle), LV with regional dyskinetic (bottom): during diastole (a), (b); and systole (c), (d) [61].
Fig. 8. KaHMo flow simulation in healthy LV (volunteer) and DCM patient before
surgery (pre-op) and after surgery (post-op) [62].
as prescribed movement for LV flow analysis. KaHMo circulation
model provided pressure boundary conditions, while fluid velocity and pressure were solved with electric resistance, inductivity
and compliance of arteries and venous by taking the rheological
properties of blood into consideration. A healthy LV and infarcted
LV before and after surgery were simulated in the study. The
infarction occurred at the anterior of the apex.
In the healthy LV, the vortex rings were observed but these
were asymmetrically formed during the filling phase. The vortex
rings expanded and rotated in a clockwise direction to redirect
intraventricular blood, including blood at the apical region,
toward the aortic outflow. The secondary vortex ring, which
was newly formed in the basal region, was flushed out, together
with the blood stream during systole (Fig. 8: volunteer). In the
infarcted LV before surgery, aneurysm in apical region caused the
deformation of the vortex where the blood branch in the middle
of the LV resulted in flow loss during diastole. The flow during
systole coincided with the flow pattern of the healthy LV (Fig. 8:
pre-op). With the healthy LV, the same topologic LV structure was
observed in the infarcted patient before surgery. However, the
flow structure of the blood after surgery was completely different.
Removing the aneurysm region from the infarcted LV produced a
ball-shaped LV geometry and this had a significant effect on
intraventricular flow (Fig. 8: post-op). During diastole, the inflow
jet did not decelerate enough and a stagnant region was found at
the apex. The change of geometry shape also restricted the ring
vortex expansion area; as such, blood was not redirected toward
the aortic outlet during systole.
The improved EF but slightly lower stroke volume obtained
from the infarcted LV after surgery (post-operation) compared to
before surgery (pre-operation). Also the exchange transfusion
parameter was evaluated corresponding to the remaining
384
B.T. Chan et al. / Computers in Biology and Medicine 43 (2013) 377–385
fractional blood volume in the LV after a number of cardiac cycles.
It was found that blood washout in the LV of post-operation was
slightly different from the pre-operation condition. Based on this
result, the ball-shaped LV showed impaired fluid dynamics but
was not deemed to be of greater risk of thrombosis formation.
4. Future direction
CFD greatly contributes to cardiac diagnosis with condition
that accuracy of simulation result is promised. Due to the
limitation of certain aspects, assumptions are always made when
the CFD method is used. As such, result validation is important,
where well-matched outcome with imaging modalities and
experiment helps in justification. In order to have exactly compatible results, the CFD boundary conditions should be as realistic
as possible. The integration of other chambers i.e. LA and RV, as
well as complex surface topography of the ventricle, such as
trabeculae and papillary muscles, would strengthen the reliability
of CFD study. For this reason, patient-specific CFD models have
been developed with conjunction use of boundary conditions
extracted from imaging technology. Cardiovascular CFD studies
have progressively stepped forward to fluid structure interaction
(FSI), which tends to provide more realistic cardiac flow analysis
with myocardial mechanics interaction. Furthermore, the material properties of blood and myocardium play crucial roles in
response to pragmatic disease states, for instance, non-Newtonian
blood and myocardial viscoelasticity properties are contemporary
representation. Ultimately, a whole cardiac model that integrates
geometrical, mechanical, electrical and hemodynamics functions
would be useful in total heart function representation [64].
After years of stability and concrete assurance, CFD should
stand in place of clinical practice. CFD provides insights into
cardiac hemodynamics and disease causation and this directly
helps in the detection of diseases during the initial stages, which
can increase survival rates. In addition, interventions of therapeutic and surgical procedures, as well as cardiac assisting
devices, are optimized through CFD trends.
5. Conclusion
To inspect flow abnormalities within a diseased heart, it is
important to have knowledge of the blood flow process and
distribution in a healthy heart [65], which is characterized by
efficient and consistent flow patterns. Previous studies have
shown that the effects of pathological LV on intraventricular flow
depend on the severity of the disease. Different pathological
patterns and affected areas have different effects on the blood
flow pattern. The DCM and MI are characterized by enlarged LV
chambers and reductions in wall motion, both globally and
regionally. The blood flow pattern deviation was shown through
vortex dynamics while systolic function was evaluated in EF.
Thrombus and its potential formation were detected in DCM and
MI through the use of the CFD method.
A prescribed geometry [59–62] and the FSI [66] method in CFD
approach were used to study hemodynamics within the LV
chamber, where the outcomes are both presented. CFD demonstrates hemodynamics evolution and is able to predict the effects
of any changes within the pathological LV. Therefore strong
comprehension in the pathophysiology of heart disease is developed and this can be used to stimulate the invention of a heartassisting device, as well as identifying methods of improving
surgical procedures.
Conflict of interest statement
I declare that we do not have any financial or personal
relationship with other people or organizations that could have
inappropriately influenced this study.
Acknowledgment
This study was funded by UM/MOHE HIR, Grant number UM.C/
HIR/MOHE/ENG/14 D000014-16001.
References
[1] Braunwald, Braunwald’s heart disease, in: P. Libby, R.O. Bonow, D.L. Mann,
D.P. Zipes (Eds.), A Textbook of Cardiovascular Medicine, Saunders Elsevier,
2007.
[2] S.J. Sarnoff, E. Berglund, Ventricular function: I. Starling’s law of the heart
studied by means of simultaneous right and left ventricular function curves
in the dog, Circulation 9 (1954) 706–718.
[3] E.H. Starling, M.B. Visscher, The regulation of the energy output of the heart,
J. Physiol. 62 (1926) 243–261.
[4] S.W. Patterson, H. Piper, E.H. Starling, The regulation of the heartbeat,
J. Physiol. 48 (1914) 357–379.
[5] H. Pouleur, M.F. Rousseau, C. Van Eyll, J. Melin, M. Youngblood,
S.A.L.I.M. Yusuf, Cardiac mechanics during development of heart failure,
SOLVD Investigators. Circulation 87 (5 Suppl.) (1993) IV14.
[6] D.L. Mann, Mechanisms and models in heart failure, Circulation 100 (1999)
999–1008.
[7] G.P. Aurigemma, M.R. Zile, W.H. Gaasch, Contractile behavior of the left
ventricle in diastolic heart failure with emphasis on regional systolic function, Circulation 113 (2006) 296–304.
[8] S.M. Mathew, et al., Diastolic dysfunction: can it be diagnosed by Doppler
echocardiography? J. Am Coll. Cardiol. 44 (2004) 1543–1549.
[9] Y. Amano, et al., MRI of cardiac morphology and function after percutaneous
transluminal septal myocardial ablation for hypertrophic obstructive cardiomyopathy, Am. J. Roentgenol. 182 (2) (2004) 523–527.
[10] J. Somauroo, et al., An echocardiographic assessment of cardiac morphology
and common ECG findings in teenage professional soccer players: reference
ranges for use in screening, Heart 85 (6) (2001) 649–654.
[11] M.I. Altbach, et al., Cardiac MRI is complementary to echocardiography in the
assessment of cardiac masses, Echocardiography 24 (3) (2007) 286–300.
[12] A. Rosen, G.T. Inouye, A.L. Morse, D.L. Judge, Magnetic recordings of the
heart’s electrical activity with a cryogenic magnetometer, J. Appl. Phys. 42
(10) (1971) 3682–3684.
[13] K.V.L. Narayana, Wavelet based QRS detection in ECG using MATLAB,
Innovative Syst. Des. Eng. 2 (7) (2011) 60–69.
[14] K.F. Augenstein, E.R. McVeigh, A.A. Yooung, Magnetic resonance imaging and
ventricle mechanics, Philos. Trans. R. Soc. Lond. A 359 (2001) 1263–1275.
[15] J.N. Kirkpatrick, R.M. Lang, Insights into myocardial mechanics in normal and
pathologic states using newer echocardiographic techniques, Curr. Heart Fail.
Rep. 5 (2008) 3.
[16] J.D. Thomas, Z.B. Popovic, Assessment of left ventricular function by cardiac
ultrasound, J. Am. Coll. Cardiol. 48 (10) (2006) 2012–2025.
[17] A. Feintuch, et al., Hemodynamics in the mouse aortic arch as assessed by
MRI, ultrasound, and numerical modeling, Am. J. Physiol. 292 (2) (2007)
H884–H892.
[18] D. Augst, et al., Accuracy and reproducibility of CFD predicted wall shear
stress using 3D ultrasound images, ASME J. Biomed. Eng. 125 (2003)
218–222.
[19] C.S. Peskin, D.M. McQueen, A three-dimensional computational method for
blood flow in the heart, J. Comput. Phys. 81 (1989) 372–405.
[20] F. Domenchini, G. Pedrizzetti, B. Baccani, Three-dimensional filling flow into
a model left ventricle, J. Fluid. Mech. 539 (2009) 179–198.
[21] J.A. Vierendeels, K. Riemslagh, E. Dick, Computer simulation of intraventricular flow and pressure gradients during diastole, J. Biomech. Eng. 122 (2000)
667–674.
[22] Q. Long, et al., Subject-specific computational simulation of left ventricular
flow based on magnetic resonance imaging, Proc. IMechE Part H: J. Eng. Med.
222 (2008) 475–485.
[23] T. Schenkel, et al., MRI-based CFD analysis of flow in a human left ventricle:
methodology and application to a healthy heart, Ann. Biomed. Eng. 37 (3)
(2008) 502–515.
[24] N.R. Saber, et al., Computational flow modeling of the left ventricle based on
in vivo MRI data: initial experience, Ann. Biomed. Eng. 29 (2001) 275–283.
[25] C.A. Taylor, C.A. Figueroa, Patient-specific modeling of cardiovascular
mechanics, Annu. Rev. Biomed. Eng. 11 (2009) 109–134.
[26] Y.G. Cheng, H. Oertel, T. Schenkel, Fluid–structure coupled CFD simulation of
the left ventrivular flow during filling phase, Ann. Biomed. Eng. 33 (5) (2005)
567–576.
B.T. Chan et al. / Computers in Biology and Medicine 43 (2013) 377–385
[27] H. Watanabe, et al., Multiphysics simulation of left ventricular filling
dynamics using fluid–structure interaction finite element method, Biophys.
J. 87 (2004) 2074–2085.
[28] D. Nordsletten, et al., Fluid–solid coupling for the investigation of diastolic
and systolic human left ventricular function, Int. J. Numer. Methods Biomed.
Eng. 27 (2011) 1017–1039.
[29] S. Krittian, et al., Partitioned fluid–solid coupling for cardiovascular blood
flow, Ann. Biomed. Eng. 38 (4) (2010) 1426–1441.
[30] J.D. Lemmon, A.P. Yoganathan, Three-dimensional computational model of
left heart diastolic function with fluid–structure interaction, J. Biomech. Eng.
122 (2000) 109–117.
[31] S.D. Solomon, et al., Influence of ejection fraction on cardiovascular outcomes
in a broad spectrum of heart failure patients, Circulation 112 (24) (2005)
3738–3744.
[32] M. Galderisi, Diastolic dysfunction and diastolic heart failure: diagnostic,
prognostic and therapeutic aspects, Cardiovasc. Ultrasound 3 (9) (2005)
1–14.
[33] J.L. Jefferies, J.A. Towbin, Dilated cardiomyopathy, Lancet 375 (9716) (2010)
752–762.
[34] H. Imai, et al., Left ventricular trabeculae evaluated with MRI in dilated
cardiomyopathy and old myocardial infarction, J. Cardiol. 22 (1) (1992)
83–90.
[35] C.S. Rihal, et al., Systolic and diastolic dysfunction in patients with clinical
diagnosis of dilated cardiomyopathy relation to symptoms and prognosis,
Circulation 90 (6) (1994) 2772–2778.
[36] D.E. Thomas, et al., The role of echocardiography in guiding management in
dilated cardiomyopathy, Eur. Heart J. 10 (8) (2009) iii15–iii21.
[37] E. Donal, et al., Mitral regurgitation in dilated cardiomyopathy: value of both
regional left ventricular contractility and dyssynchrony, Eur. J. Echocardiogr.
10 (2009) 133–138.
[38] M.A. Fisher, M.T. McNamara, C.B. Higgins, Acute myocardial infarction: MR
evaluation in 29 patients, Am. J. Roentgenol. 148 (1986) 247–251.
[39] A. de Roos, et al., MR imaging of acute myocardial infarction: value of
Gd-DTPA, Am. J. Roentgenol. 150 (1988) 531–534.
[40] F.M. Baer, et al., Chronic myocardial infarction: assessment of morphology,
function, and perfusion by gradient echo magnetic resonance imaging and
99mTc-methoxyisobutyl-isonitrile SPECT, Am. Heart J. 123 (3) (1992)
636–645.
[41] J.V. Nixon, K.A. Narahara, T.C. Smitherman, Estimation of myocardial involvement in patients with acute myocardial infarction by two-dimensional
echocardiography, Circulation 62 (1980) 1248–1255.
[42] J. Heikkila, M. Nieminen, Echoventriculographic detection, localization, and
quantification of left ventricular asynergy in acute myocardial infarction: a
correlative echo- and electrocardiographic study, Br. Heart J. 37 (1975)
46–59.
[43] S. Wilansky, Echocardiography in the assessment of complications of myocardial infarction, Tex. Heart Inst. J. 18 (1991) 237–242.
[44] B.M. Jackson, et al., Extension of borderzone myocardium in postinfarction
dilated cardiomyopathy, J. Am. Coll. Cardiol. 40 (2002) 1160–1167.
[45] S. Beppu, et al., Abnormal blood pathways in left ventricular cavity in acute
myocardial infarction. Experimental observations with special reference to
regional wall motion abnormality and hemostasis, Circulation 78 (1988)
157–164.
[46] M. Thomas, R. Malmcrona, J. Shillingford, Hemodynamic changes in patients
with acute: myocardial infarction, Circulation 31 (6) (1965) 811–823.
[47] J.B. Dahm, et al., Doppler echocardiography in dilated cardiomyopathy:
diastolic and combined systolic/diastolic parameters offer more detailed
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]
[56]
[57]
[58]
[59]
[60]
[61]
[62]
[63]
[64]
[65]
[66]
385
information on left ventricular global dysfunction than systolic parameters,
J. Clin. Basic Cardiol. 5 (2) (2002) 189–192.
H. Persson, et al., Heart failure after myocardial infarction: the importance of
diastolic dysfunction: a prospective clinical and echocardiographic study,
Eur. Heart J. 16 (4) (1995) 496–505.
A.M. Shutov, N.I. Kondratyeva, E.S. Kulikova, Does E/A ratio correctly estimate
diastolic dysfunction in patients with chronic renal failure? Nephrol. Dial.
Transplant. 15 (8) (2000) 1256–1257.
S.W. Lee, et al., E/E’ ratio is more sensitive than E/A ratio for detection of left
ventricular diastolic dysfunction in systemic lupus erythematosus, Lupus 17
(3) (2008) 195–201.
K. Hu, et al., Failure to unmask pseudonormal diastolic function by a valsalva
maneuver: tricuspid insufficiency is a major factor, Cir.: Cardiovas. Imaging 4
(6) (2011) 671.
Jithendra B. Somaratne, et al., Pseudonormal mitral filling is associated with
similarly poor prognosis as restrictive filling in patients with heart failure
and coronary heart disease: a systematic review and meta-analysis of
prospective studies, Prog. Cardiovasc. Dis. 22 (5) (2009) 494–498.
E. Castillo, J.C. Lima, D.A. Bluemke, Regional myocardial function: advances in
MR imaging and analysis, RadioGraphics 23 (2003) S127–S140.
M. Discacciati, A. Quarteroni, Navier–Stokes/Darcy coupling: modeling,
analysis, and numerical approximation, Rev. Mat. Comput. 22 (2) (2009)
315–426.
J.R. Nanduri, F.A. Pino-Romainville, I. Celik, CFD mesh generation for biological flows: geometry reconstruction using diagnostic images, Comput. Fluids
38 (2009) 1026–1032.
C. Taylor, M. Draney, Experimental and computational methods in cardiovascular fluid mechanics, Ann. Rev. Fluid Mech. 36 (2004) 197–231.
J. Del Alamo, Recent advances in the application of computational mechanics
to the diagnosis and treatment of cardiovascular disease, Rev. Esp. Cardiol. 62
(7) (2009) 781–805.
L. Ge, M. Ratcliffe, The use of computational flow modeling (CFD) to
determine the effect of left ventricular shape on blood flow in the left
ventricle, Ann. Thorac. Surg. 87 (2009) 993–994.
B. Baccani, et al., Fluid dynamics of the left ventricular filling in dilated
cardiomyopathy, J. Biomech. 35 (2002) 665–671.
S. Loerakker, et al., Influence of dilated cardiomyopathy and a left ventricular
assist device on vortex dynamics in the left ventricle, Comput. Methods
Biomech. Biomed. Eng. 11 (6) (2008) 649–660.
F. Domenichini, G. Pedrizzetti, Intraventricular vortex flow changes in the
infarcted left ventricle: numerical results in an idealised 3D shape, Comput.
Methods Biomech. Biomed. Eng. 14 (1) (2011) 95–101.
T. Doenst, et al., Fluid-dynamic modeling of the human left ventricle:
methodology and application to surgical ventricular reconstruction, Ann.
Thorac. Surg. 87 (2009).
T. Schenkel, et al. The Karlsruhe Heart Model Kahmo: A Modular Framework
for Numerical Simulation of Cardiac Hemodynamics in IFMBE Proceedings,
2009.
M. Sermesant, K. Rhode, G.I. Sanchez-Ortiz, O. Camara, R. Andriantsimiavona,
S. Hegde, D. Rueckert, et al., Simulation of cardiac pathologies using an
electromechanical biventricular model and XMR interventional imaging,
Med. Image Anal. 9 (2005) 467–480.
C.J. Carlhall, A. Bolger, Passing strange: flow in the failing ventricle, Circulation 3 (2010) 326–331.
H. Watanabe, et al., Finite element analysis of ventricular wall motion and
intraventricular blood flow in heart with myocardial infarction, JSME Int. J.
47 (4) (2004) 1019–1026.