Você está na página 1de 11

Annals of Biomedical Engineering, Vol. 36, No. 11, November 2008 ( 2008) pp.

1753–1763
DOI: 10.1007/s10439-008-9558-0

Hemodynamic Analysis of a Compliant Femoral Artery Bifurcation Model


using a Fluid Structure Interaction Framework
YOUNG-HO KIM,1 JONG-EUN KIM,1 YASUSHI ITO,1 ALAN M. SHIH,1 BRIGITTA BROTT,2
and ANDREAS ANAYIOTOS3,4
1
Department of Mechanical Engineering, University of Alabama at Birmingham, Birmingham, AL, USA; 2Division of
Cardiovascular Disease, University of Alabama at Birmingham, Birmingham, AL, USA; 3Department of Biomedical
Engineering, University of Alabama at Birmingham, 1075 13th Street South, Hoehn 368, Birmingham, AL 35294-4440, USA;
and 4Department of Mechanical and Materials Engineering, Cyprus University of Technology, Limassol 3036, Cyprus
(Received 25 May 2007; accepted 26 August 2008; published online 16 September 2008)

Abstract—The influence of wall motion on the hemodynamic INTRODUCTION


characteristics of the human femoral bifurcation and its
effects on the development of peripheral artery disease has Lower limb peripheral arterial disease (PAD) affects
not been previously investigated. This study aimed in approximately 8 million Americans.19 It is particularly
investigating the hemodynamics of a compliant patient-
prevalent in elderly persons, affecting 12–20% those
specific femoral artery bifurcation model by a fluid structure
interaction (FSI) scheme. The complex physiological geom- over 65 years of age.38,43 Even in those who are
etry of the femoral artery bifurcation was reproduced from asymptomatic, it is associated with impaired function
sequentially obtained transverse CT scan images. Velocity and quality of life.34 Atherosclerosis is widely accepted
waveforms derived from phase contrast MR images were as the main causative factor in the development of
extracted and mapped to define boundary conditions. Equa-
lower limb PAD. The atherosclerosis in PAD patients
tions governing blood flow and wall motion were solved
using an FSI framework that utilizes commercial codes: is frequently observed on the site at the femoral artery
FLUENT for computational fluid dynamics and ANSYS for bifurcation at the division of the Common Femoral
computational structural dynamics. The results showed that Artery (CFA) into the Superficial Femoral Artery
wall compliance decreased flow velocities at the relatively (SFA) and the Profunda Femoris Artery (PFA).14,15
high curvature geometries including common and superficial
The PFA provides blood to the thigh muscle, while
femoral artery (SFA), and it created strong recirculation in
the profunda femoris artery close to the bifurcation. In the the SFA provides blood to the lower limb and extends
SFA region near the apex, time averaged wall shear stress to the popliteal artery. The lateral circumflex artery
(TAWSS) differences up to 25% between compliant and rigid (LCA), branches out from the PFA, the location of
models were observed. The compliant model also exhibited which may vary in each patient. The development of
lower TAWSS and oscillatory shear at the superior section of
atherosclerotic plaque in the femoral artery restricts
the common femoral artery close to the bifurcation. The
presence of wall motion, however, created minor differences blood supply to the lower extremities, leading to
in the general flow-field characteristics. We conclude that ischemia and lower limb disability. The distensibility of
wall motion does not have significant influence on the global the arterial wall has usually been neglected in arterial
fluid dynamic characteristics of the femoral artery bifurca- hemodynamic studies. Wall motion, however, may
tion. Longer arterial segments need to be simulated to see the
influence the distribution of wall shear stress (WSS),
effect of wall motion on tortuousity which was previously
cited as an important factor in the development of athero- oscillatory shear index (OSI), wall shear stress tem-
sclerosis at the femoral artery. poral gradient (WSSTG), curvature, and wall rough-
ness, which have previously been cited as having
Keywords—Computational fluid dynamics, Fluid structure an important role in the initiating mechanisms of
interaction, Blood flow, Wall shear stress, CT image-based atherosclerosis.3,7,44–46,48,52
model. Experimental studies on arterial segments including
femoral artery bifurcations,7 femoral graft-artery
junctions,27 carotid artery bifurcations,6,30,36 and
arteriovenous anastomoses31 using idealized or sim-
Address correspondence to Andreas Anayiotos, Department of plified arterial geometries have shown that hemody-
Biomedical Engineering, University of Alabama at Birmingham, namic characteristics are influenced by anatomic
1075 13th Street South, Hoehn 368, Birmingham, AL 35294-4440,
geometry, pulsatile flow, arterial curvature, arterial
USA. Electronic mail: aanayiot@eng.uab.edu
1753
0090-6964/08/1100-1753/0  2008 Biomedical Engineering Society
1754 KIM et al.

wall motion and arterial stenosis.1–3,6–8,16,22,26,27,31,32


Recent improvements in medical imaging techniques,
and three-dimensional geometric reconstruction of
arteries, allow computational fluid dynamic CFD
methods to investigate complex vascular flows through
patient-specific artery models.17,49,57
CFD studies, however, do not usually consider the
compliance of arterial wall due to the complexity of the
numerical modeling and the requirement for a fluid
structure interaction (FSI) environment to solve for the
coupled fluid/structure problem. The assumption of a
rigid wall in CFD analysis of arterial models may
underestimate the temporal and spatial magnitude
of slow moving zones.58 To assess the effect of wall
compliance on the near wall hemodynamic characteris-
tics of arteries such as WSS, OSI, and WSSTG, FSI
algorithms have been employed to take into account
the effect of arterial wall compliance in the fluid
dynamic analysis.10,13,29,33,56 Various arterial sites have
been analyzed by FSI such as the carotid arterial
bifurcation,54 an abdominal aortic aneurysm model,39 FIGURE 1. Structural (coarse) and fluid (fine) mesh of the
femoral artery bifurcation (A: Anterior, P: Posterior, L: Left, R:
and a stented aortic aneurysm model.28 Right views).
In this study, we aimed in assessing WSS, OSI, and
WSSTG in the femoral artery bifurcation by a newly
developed fluid structure interaction framework that
simultaneously computes both hemodynamic and used to create high-quality surface meshes. A tetra-
mechanical characteristics by taking into account the hedral mesh for CFD simulations was created by
interaction between blood flow and arterial wall subdividing non-simplicial elements in a hybrid mesh
motion. to estimate WSS precisely.20
The wall thickness of the CFA was set to 0.74 mm
by assuming a radius-to-thickness ratio of 4.122 using
the reconstructed geometry. This was adopted from a
MATERIAL AND METHODS previous reported study.48 The wall thickness could
not be obtained from the CT information due to the
Model Reconstruction and Mesh Generation
relatively large pixel size and low signal to noise ratio
A 6 cm-long biofidelic femoral bifurcation model (SNR) that did not allow clear distinction between
was reconstructed from CT data which provides a arterial wall and adjacent tissue structure. The same
better contrast of the arterial wall and lumen cross radius-to-thickness ratio was used to determine the
section than MR, but does not have the capability of thicknesses of the SFA, PFA and LCA. Figure 1
rendering the velocity measurements. A 992 CT scan illustrates structural and fluid meshes of the patient-
image set was obtained from a 55-year-old female specific femoral artery used in this study. A common
volunteer suffering from paraplegia. This dataset interface was defined at the inner wall of the femoral
ranges from the abdomen to the toes, using a GE artery. The finer mesh with 367,315 tetrahedra covers
LightSpeed Proscanner with 1.25 mm slice thickness the inside of the femoral artery and is used for flow
and 0.25 mm overlap. An informed written consent field analyses. There are 18,695 triangles on the com-
from the patient was received, and the protocol was mon interface. The coarse tetrahedral mesh repre-
approved by the local institutional review board. The senting the wall of the femoral artery is used for
60 slices of the images were used to reconstruct the structural analyses. It has 43,774 tetrahedra and 4761
femoral model. The image resolution was 512 9 512 triangles on the common interface.
pixels per slice, and each pixel represented a 0.79 9
0.79 mm2 area. Image segmentation and geometry
Fluid Structure Interaction Framework
extraction from the images were performed to obtain
the inner surface of the femoral artery. Two surface An in-house loosely coupled fluid structure inter-
triangulation methods, a direct advancing front action (FSI) framework was developed and employed
method and a modified decimation method,21 were for this study. This framework enables to select
Hemodynamic Analysis of Femoral Artery Bifurcation 1755

any well-proven CFD and computational structural encoding velocity was 150 cm/s corresponding to 180
dynamic (CSD) codes. An accurate and efficient data phase shift. Only axial velocity was measured under the
transfer between fluid and structure using interpolation assumption that in-plane velocity components are rel-
techniques is essential in space due to naturally atively trivial. The image has 0.59 mm in-plane reso-
unmatched meshes between fluid and structure. The lution. To obtain velocity map from MRI during the
framework consists of: a conservative point-wise cardiac cycles, a region of interest was determined
interpolation12 module for blood pressure transfer to manually around the vessel lumen detected age. The
wall, a curvature-gradient index local fitting method24 phase angle value in each pixel was converted to
for wall motion transfer to fluid domain, a semi- velocity by18:
torsional spring analogy55 for fluid volume mesh
V ¼ VENC  Uv =180 ; Uv ¼ ½180 ; 180  ð1Þ
movement, and a serial staggered method11 with sub-
cycling for synchronizing time levels between CFD and where, VENC is encoding velocity, and Fv is the phase
CSD solvers. The entire FSI process is automatically angle obtained by phase image. The calculated velocity
executed using Bash/Perl script files for data extrac- then was mapped by an interpolation method based on
tion, manipulation, file handling, and code execution. a radial basis function. The thin plate spline method9
The details of the FSI framework are described in the was used for the interpolation. Time-dependent
literature.23 Two commercial codes, FLUENT for velocity profiles were mapped through 32 cardiac
CFD and ANSYS for CSD, were used in this study. In phases. Around 40 pixels for the SFA and PFA, and
order to take into account arterial dynamic wall mo- 20 pixels for the LCA were used to quantify the
tion, a time integration option in ANSYS was switched velocity.
on to include the inertial effect of the wall in the FSI The mapped and interpolated flow rates used in
simulation. calculating velocity at the SFA, PFA and LCA outlets
were consistent with MRI data through the entire
cardiac cycle as shown in Fig. 3. These time-varying
MRI Velocity Mapping and Boundary Conditions
mapped flow rate were used for three outlet boundary
Phase contrast MRI data were acquired to measure conditions in FLUENT via user defined function
blood flow velocity at four different sections shown in (UDF).
Fig. 2 (39.2 mm, 19.2 mm above branch and 11 mm, CFD analysis in rigid wall vessel simulations
21 mm below branch) by a GE MR scanner (1.5 T, may use only flow rate boundary conditions. However
Signa). The scan parameters were 4 mm slice thickness, in compliant wall vessel FSI simulations, a time-
40 ms TR, 6 ms TE, 32 cardiac phases, 20 FA, dependent pressure profile, is necessary to apply forces
113 9 150 cm FOV, and a 512 9 512 matrix. The on the wall. The time dependent pressure profile at
Section 2 in Fig. 2 was used as the inlet boundary
condition. To calculate the pressure profile at Section 2,

FIGURE 3. Flow rate comparison between MRI-PCA calcu-


FIGURE 2. The four sections where the velocity profiles lated measurements and CFD simulated measurements at the
were measured by MRI-PCA. SFA, PFA and LCA outlets.
1756 KIM et al.

TABLE 1. Modified coefficients of five-coefficient Mooney-


Rivlin model for patient-specific arterial properties.

Parameters C10 C01 C20 C11 C02

Values (kPa) 4.10 2.75 590.42 857.18 0

The strain energy potential W is expressed by five


parameters Cij ‘s in incompressible Mooney–Rivlin
model:
X
2
W¼ Cij ðI1  3Þi ðI2  3Þj i; j ¼ 0; 1; 2 ð4Þ
iþj¼1

where I1 and I2 are strain invariants. In this study, the


five parameters in Eq. (4) were based on the results of
FIGURE 4. Calculated pressure at Section 2 using the
Navier–Stokes equations for one dimensional flow in an in vitro tensile test from literature.41 The value of C10,
elastic tube with flow rates based on MRI-PCA. however, was modified to consider in vitro property of
the PAD patients, in which the blood vessel is pre-
pressurized at the diastolic phase. For this, the pres-
two physiological flow rates at Sections 1 and 2 were sure difference between systolic and diastolic phase
calculated based on the velocity profiles. The Navier– from Eq. (2) was applied to the ideal cylinder. The
Stokes equations for one dimensional flow in an elastic value of C10 was determined until the diameter change
tube can be reduced to5 reached to the 6.2% diameter change of the CFA for
Qi ðtÞ  Qo ðtÞ 3pr3 dPo ðtÞ the PAD patient at a systolic phase.50 The five
¼ ð2Þ parameters used in Mooney–Rivlin model are shown
Dz 2hE dt
in Table 1.
Po ðtÞ  Pi ðtÞ q dQi ðtÞ 8l
¼ 2 þ 4 Qi ðtÞ ð3Þ
Dz pr dt pr Time Averaged WSS, OSI, and WSSTG
where Qi and Qo are the flow rates at the Sections 1 and Time averaged wall shear stress (TAWSS) was
2, respectively, Dz is the distance between Sections 1 measured by calculating temporal mean of WSS during
and 2, Pi and Po are the pressure at the Sections 1 and an entire cardiac cycle. OSI was obtained using the
2, respectively, r is the inside radius of tube, h is the form26:
wall thickness, q is the density of blood, l is dynamic RT 
viscosity of blood and E is the elastic modulus set to j~
s jdt
R 0T ð5Þ
1.61e06 Pa.37 The pressure waveform estimated by 0 j~ sx jdt
integrating Eq. (2) is illustrated in Fig. 4. This pressure
waveform was used as an inlet boundary condition at where T is the time for a cardiac cycle, ~ sx is the
the Section 2 by UDF in FLUENT. The mapped MRI instantaneous WSS vector and ~ s is the WSS compo-
velocities at Section 4 (of the PFA, SFA, and LCA), nent acting in the direction opposite to the temporal
were used as the outlet boundary conditions. mean WSS vector. The WSSTG is defined as a maxi-
mum among the gradients of WSS in two consecutive
time steps during the cycle.54
Vessel Material Properties
The blood vessel was assumed as a homogenous,
hyperelastic, nonlinear, and nearly incompressible RESULTS
material. Poisson’s ratio was set to 0.4999 to depict
Wall Displacement
incompressible material. The density of the vessel was
set to 1100 kg/m3. The in vivo material properties of Time-varying lumen diameters of the CFA and SFA
the CFA were identified through inverse finite element are measured at Sections 2 and 3 as shown in Fig. 2,
(FE) analysis. An ideal cylinder FE model that surro- respectively. The maximum anterior-to-posterior
gates the vessel from Section 1 to 2 in Fig. 2 was lumen diameters of CFA and SFA at peak systole are
generated. Mooney–Rivlin hyperelastic model35 was 6.35 and 4.67 mm, respectively. This corresponds to
used for the constitutive model of compliant arteries. 5.7% and 3.7% expansion of the lumen diameter at
Hemodynamic Analysis of Femoral Artery Bifurcation 1757

FIGURE 6. Time-dependent velocity at Section 3: (a) Near


right wall; (b) Near posterior wall.
FIGURE 5. Time-dependent lumen diameter changes: (a)
CFA at Sections 2; (b) SFA at Sections 3.
observed. The recirculation is more accentuated later
in diastole for phase 0.54/0.78 s. The velocity contours
in Fig. 8 show that wall compliance decreases flow
peak systole as shown in Fig. 5. The increase of ante- velocities at the posterior CFA at Section 2 and the
rior-to-posterior lumen diameters of CFA and SFA is anterior SFA at Section 3 where the radius of curva-
larger than that of right-to-left ones. ture is relatively high.

Flow Pattern Comparison Wall Shear Stress


The validity of the CFD and FSI models in com- Comparison of the instantaneous peak systolic WSS
parison to the raw MRI data in the model, are dem- distributions between the compliant and rigid cases is
onstrated in the time histories of Figs. 6a and 6b, shown in Fig. 9. A maximum difference between the
where good agreement is seen. The hemodynamic two models is shown at the left anterior view close to
characteristics of the compliant model through the FSI the branch of PFA. The plots demonstrating the dif-
analysis are compared to those of the rigid wall model ference in TAWSS, WSSTG, and OSI between the two
using CFD analysis. Figures 7a and 7b show the cases are illustrated in Figs. 10–12, respectively. As
instantaneous streamlines during the deceleration seen in Fig. 9, high WSS is shown at the branch of
phase of the cycle at phases 0.23/0.78 s and 0.54/0.78 s, PFA and bifurcation. Low WSS regions appear
respectively. More recirculation in the PFA section around the outer wall of the posterior view upstream
close to the bifurcation in the compliant model is of the bifurcation and on anterior to left aspect close to
1758 KIM et al.

FIGURE 7. Comparison of streamlines between rigid and compliant arteries: (a) at time 0.2926/0.78 s; (b) at time 0.5368/0.78 s.

the branch of PFA where the geometry has a mild that there are significant differences up to 25% in wall
curvature along the distal direction. In these regions, shear stress temporal gradient (WSSTG) between a
the TAWSS for the compliant model appears to be rigid and a compliant model of the femoral artery. The
only 10% lower than the rigid model, and there are 10–25% difference regions (Fig. 11) cover a wide
some patchy regions that illustrate OSI differences up extent in the superficial femoral artery near the bifur-
to 25%. In contrast, the WSSTG difference contour cation. There is also a significant region of differences
plots in Figs. 11a and 11b show the highest (25%) from 5 to 20% at a region of the common femoral
difference on the posterior right side of the SFA near artery near the bifurcation (Fig. 11). Our results also
the bifurcation along the distal direction. show that there are some regions of differences in
oscillatory shear index (OSI) that reach up to 25% at
the common femoral artery close to the bifurcation.
DISCUSSION There has not been a previous study on the com-
pliance of the femoral bifurcation, however, Younis
Correct modeling of fluid dynamics and the char- and colleagues54 used an FSI technique to study an
acterization of the near wall phenomena is important image based model of the carotid bifurcation. They
in understanding the initializing mechanisms and the reported minor differences in TAWSS differences
development of atherosclerotic plaque as well as pre- between the rigid and compliant models. They found,
dicting changes that may occur with treatments such as however, significant differences in the WSSTG and
stent placement or surgical repair. Our results indicate OSI between the two models except at the apex. In our
Hemodynamic Analysis of Femoral Artery Bifurcation 1759

FIGURE 8. Comparison of velocity between rigid and compliant arteries at peak pressure loads: (a) Rigid CFA at Section 2;
(b) Compliant CFA at Section 2; (c) Rigid SFA at Section 3; Compliant SFA at Section 3.

study, the wall compliance created minor TAWSS and WSSTG, curvature, wall roughness and wall
differences at most locations, but up to 25% differ- motion have also been indicated to have a role in the
ences in WSSTG and OSI at some locations. We had initializing mechanisms of atherosclerosis at the fem-
observed 20% reduction in instantaneous WSS for the oral artery bifurcation.7,44–46,52 The current study
compliant model around the branch at peak systole could not assess the effects of wall motion in changes
which is somewhat lower than in what is reported in in roughness, or curvature.
previous studies.2,40 However, this can be explained by Earlier experimental studies have also revealed that
the lower distensibility of the muscular femoral arter- the inherent tortuousity of the femoral artery increases
ies which was modeled as opposed to greater disten- the rate of atherosclerosis and that the inner curvature
sibility of large elastic arteries such as aortic and walls are more prone to atherosclerotic plaques than
carotid.25,42,53 the outer curvature walls.45,46 These studies also sug-
It has been widely accepted that WSS and related gested that the curvature of the SFA is more of a risk
parameters such as WSSTG and OSI are important factor for atherosclerosis than the presence and loca-
factors in the development of atherosclerosis on the tion of branches. Wood and colleagues52 showed a
femoral artery,22,46,47,51 and the carotid artery.1,2,16,26 strong correlation between tortuousity and flow dis-
Ku and colleagues26 reported low OSI in addition turbance in computational reconstructed SFA’s. They
to low WSS enhanced the development of plaques. suggested that the regions of low time averaged WSS
Bao and colleagues4 showed that WSSTG leads to might be high-risk regions of atherogenesis and pro-
enhanced endothelial proliferation. gression of atheroma. A limitation of this present study
Our data show that compliance increases the regions is that the arterial segment simulated had significant
of high OSI and WSSTG making thus specific regions curvature but no tortuousity. Longer sections of
of the femoral artery bifurcation more prone to the this artery may be able to incorporate the effect of
development of disease. However, in addition to OSI tortuousity and need to be simulated in the future to
1760 KIM et al.

FIGURE 9. Instantaneous WSS contours of rigid and compliant artery and difference contour between them at peak systole and at
different angles: (a) WSS of rigid artery at right; (b) WSS of rigid artery at left; (c) WSS of compliant artery at right; (d) WSS of
compliant artery at left; (e) WSS difference at right; (f) WSS difference at left.

evaluate the effect of wall compliance on tortuousity differences of 25% in WSSTG were observed at the
WSSTG and OSI. bifurcation on the side of the SFA. The general
velocity field was essentially insensitive to wall motion.
CONCLUSIONS We conclude that wall compliance at the femoral
artery bifurcation does not significantly change
Using an in house FSI system, the flow character- parameters previously reported as important in the
istics and WSS field of a compliant femoral artery development of arterial disease but longer arterial
bifurcation model was studied in comparison to a rigid segments need to be simulated to see the effect of wall
model. Minor differences were found in TAWSS and motion on tortuousity.
Hemodynamic Analysis of Femoral Artery Bifurcation 1761

FIGURE 10. TAWSS difference contour between them at different angles: (a) TAWSS difference at right; (b) TAWSS difference at
left.

FIGURE 11. WSSTG difference contour between them at different angles: (a) WSSTG difference at right; (b) WSSTG difference at
left.

FIGURE 12. OSI difference contour between them at different angles: (a) OSI difference at right; (b) OSI difference at left.
1762 KIM et al.
13
ACKNOWLEDGMENT Friedman, M. H., G. M. Hutchins, C. B. Bargeron, O. J.
Deters, and F. F. Mark. Correlation between intimal
The authors would like to thank Mr. Corey Shum in thickness and fluid shear in human arteries. Atherosclerosis
the Enabling Technology Laboratory, University of 39(3):425–436, 1981. doi:10.1016/0021-9150(81)90027-7.
14
Alabama at Birmingham for extracting geometry data Futterman, L. G., and L. Lemberg. Peripheral arterial
disease is only the tip of the atherosclerotic ‘‘Iceberg’’. Am.
from the CT scan.
J. Crit. Care. 11(4):390–394, 2002.
15
Garasic, J. M., and M. A. Creager. Percutaneous inter-
ventions for lower-extremity peripheral atherosclerotic
disease. Rev. Cardiovasc. Med. 2(3):120–125, 2001.
REFERENCES 16
Gijsen, F. J. H., D. E. M. Palmen, M. H. E. van der Beek, F.
N. van de Vosse, M. E. H. van Dongen, and J. D. Janssen.
1
Anayiotos, A., S. Jones, D. P. Giddens, S. Glagov, and C. Analysis of the axial flow field in stenosed carotid artery
K. Zarins. Shear stress at a compliant model of the human bifurcation models-LDA experiments. J. Biomech.
carotid bifurcation. J. Biomech. Eng. 116(1):98–106, 1994. 29(11):1483–1489, 1996. doi:10.1016/0021-9290(96)84544-1.
17
doi:10.1115/1.2895710. Giordana, S., S. J. Sherwin, J. Peiro, D. J. Doorly, J. S.
2
Anayiotos, A., P. D. Pedroso, E. C. Elefhteriou, Crane, K. E. Lee, N. J. W. Cheshire, and C. G. Caro. Local
R. Venugopalan, and W. Holman. Effect of a flow- and global geometric influence on steady flow in distal
streamlining implant at the distal anastomosis of a coro- anastomoses of peripheral bypass grafts. J. Biomech. Eng.
nary artery bypass graft. Ann. Biomed. 30(7):917–926, 127(7):1087–098, 2005. doi:10.1115/1.2073507.
18
2002. doi:10.1114/1.1500407. Higgins, C. B., and A. Roos. MRI and CT of the
3
Back, L. H., M. R. Back, E. Y. Kwack, and D. W. Cardiovascular System, 2nd edn. Lippincott Williams &
Crawford. Flow measurements in a human femoral artery Wilkins, 2005.
19
model with reverse lumen curvature. J. Biomech. Eng. Hirsch, A. T., M. H. Criqui, D. T. Jacobson, J. G.
110(4):300–309, 1988. Regensteiner, M. A. Creager, J. W. Olin, S. H. Krook, D.
4
Bao, X., C. Lu, and J. A. Frangos. Temporal gradient in B. Hunninghake, A. J. Comerota, M. E. Walsh, M. M.
shear but not steady shear stress induces PDGF-A and McDermott, and W. R. Hiatt. Peripheral arterial disease
MCP-1 expression in endothelial cells: role of NO, NF detection, awareness, and treatment in primary care.
kappa B, and egr-1. Arterioscler. Thromb. Vasc. Biol. JAMA 286(11):1317–1324, 2001. doi:10.1001/jama.286.11.
19(4):996–1003, 1999. 1317.
20
5
Flow in large blood vessels, Fluid Dynamics in Biology, Ito, Y., A. M. Shih, B. K. Soni, and K. Nakahashi.
Contemporary Math. Series, edited by A. Y. Cheer and C. P. Multiple marching direction approach to generate high
Van Dam, Amer. Math. Soc. Providence, 1992, pp. 479–518. quality hybrid meshes. AIAA J. 45(1):162–167, 2007.
6
Botnar, R., G. Rappitsch, M. B. Scheidegger, D. Liepsch, doi:10.2514/1.23260.
21
K. Perktold, and P. Boesiger. Hemodynamics in the car- Ito, Y., P. C. Shum, A. M. Shih, B. K. Soni, and K. Na-
otid artery bifurcation: a comparison between numerical kahashi. Robust generation of high-quality unstructured
simulations and in vitro MRI measurements. J. Biomech. meshes on realistic biomedical geometry. Int. J. Numer.
33(2):137–144, 2000. doi:10.1016/S0021-9290(99)00164-5. Meth. Eng. 65(6):943–973, 2006. doi:10.1002/nme.1482.
22
7
Cho, Y. I., L. H. Back, and D. W. Crawford. Pressure Jones, C. J. H., M. J. Lever, Y. Ogasawara, K. H. Parker,
difference-flow rate variation in a femoral artery branch K. Tsujioka, O. Hiramatsu, K. Mito, C. G. Caro, and F.
casting of man for steady flow. J. Biomech. Eng. Kajiya. Blood velocity distributions within intact canine
105(3):258–262, 1983. arterial bifurcations. Am. J. P. 262:H1592–H1599, 1992.
23
8
Cho, Y. I., L. H. Back, and D. W. Crawford. Experimental Kim, Y. H. Development of efficient algorithms for fluid-
investigation of branch flow ratio, angle, and Reynolds structure interaction framework and its applications. Ph.D.
number effects on the pressure and flow fields in arterial Dissertation, University of Alabama at Birmingham, 2006.
24
branch models. J. Biomech. Eng. 107(3):257–267, 1985. Kim, Y. H., and J. E. Kim. New hybrid interpolation
9
Duchon, J. Splines minimizing rotation-invariant semi- method for motion transfer in fluid structure interactions.
norms in Sobolev spaces. Constructive Theory of Func- J. Aircraft 43(2):567–569, 2006. doi:10.2514/1.15097.
25
tions of Several Variables. Lect. Notes Math. 571:85–100, Kornet, L., A. P. Hoeks, J. Lambregts, and R. S. Reneman.
1977. Mean wall shear stress in the femoral arterial bifurcation is
10
Duncan, D. D., C. B. Bargercon, S. E. Borchart, O. J. low and independent of age at rest. J. Vasc. Res. 37(2):
Deters, S. A. Gearhart, F. F. Mark, and M. H. Friedman. 112–122, 2000. doi:10.1159/000025722.
26
The effect of compliance on wall shear in casts of a human Ku, D. N., D. P. Giddens, C. K. Zarins, and S. Glagov.
aortic bifurcation. J. Biomech. Eng. 112(2):183–188, 1990. Pulsatile flow and atherosclerosis in the human carotid
doi:10.1115/1.2891170. bifurcation. Arteriosclerosis 5(3):293–302, 1985.
27
11
Farhat, C., and M. Lesoinne. Two efficient staggered Lei, M., C. Kleinstreuer, and J. P. Archie. Geometric
algorithms for the serial and parallel solution of three- design improvements for femoral graft-artery junctions
dimensional nonlinear transient aeroelastic problems. mitigating restenosis. J. Biomech. 29(12):1605–1614, 1996.
28
Comput. Meth. 182:499–515, 2000. Li, Z., and C. Kleinstreuer. Fluid-structure interaction
12
Farhat, C., M. Lesoinne, and P. LeTallec. Load and mo- effects on sac-blood pressure and wall stress in a stented
tion transfer algorithms for fluid/structure interaction aneurysm. J. Biomech. Eng. 127(4):662–671, 2005.
problems with non-matching discrete interfaces: momen- doi:10.1115/1.1934040.
29
tum and energy conservation, optimal discretization and Liepsch, D., and S. Moravec. Pulsatile flow of non-
application to aeroelasticity. Comput. Meth. 157:95–114, newtonian fluid in distensible models of human arteries.
1998. Biorheology 21(4):571–586, 1984.
Hemodynamic Analysis of Femoral Artery Bifurcation 1763
30
LoGerfo, F. W., M. D. Nowak, W. C. Quist, H. M. effect? Ann. Biomed. 24(4):474–480, 1996. doi:10.1007/
Crawshaw, and B. K. Bharadvaj. Flow studies in a model BF02648109.
46
carotid bifurcation. Arteriosclerosis 1(4):235–241, 1981. Smedby, O., J. Johansson, J. Molgaard, A. G. Olsson, G.
31
Loth, F., N. Arslan, P. F. Fischer, C. D. Bertram, S. E. Lee, Walldius, and U. Erikson. Predilection of atherosclerosis
T. J. Royston, R. H. Song, W. E. Shaalan, and H. S. for the inner curvature in the femoral artery. A digitized
Bassiouny. Transitional flow at the venous anastomosis of an angiography study. Arterioscler. Thromb. Vasc. Biol.
arteriovenous graft: Potential relationship with activation of 15(7):912–917, 1995.
47
the ERK1/2 mechanotransduction pathway. J. Biomech. Smedby, O., S. Nilsson, and L. Bergstrand. Development
Eng. 125(1):49–61, 2003. doi:10.1115/1.1537737. of femoral atherosclerosis in relation to flow disturbance.
32
Mark, F. F., C. B. Bergeron, O. J. Deters, and M. H. J. Biomech. 29(4):543–547, 1996. doi:10.1016/0021-9290
Friedman. Variations in geometry and shear rate distribu- (95)00070-4.
48
tions in casts of human aortic bifurcations. J. Biomech. Smilde, T. J., F.W. van den Berkmortel, G. H. Boers, H.
22(6–7):577–582, 1989. doi:10.1016/0021-9290(89)90009-2. Wollersheim, T. de Boo, H. van Langen, and A. F.
33
Maurits, N. M., G. E. Loots, and A. E. P. Veldman. The Stalenhoef. Carotid and femoral artery wall thickness and
influence of vessel wall elasticity and peripheral resistance stiffness in patients at risk for cardiovascular disease, with
on the carotid artery flow wave form: a CFD model com- special emphasis on hyperhomocysteinemia. Arterioscler.
pared to in vivo ultrasound measurements. J. Biomech. Thromb. Vasc. Biol. 18(12):1958–1963, 1998.
49
40(2):427–436, 2007. doi:10.1016/j.jbiomech.2005.12.008. Steinman, D. A. Image-based computational fluid dynam-
34
McDermott, M. M., K. Liu, P. Greenland, J. M. Guralnik, ics modeling in realistic arterial geometries. Ann. Biomed.
M. H. Criqui, C. Chan, W. H. Pearce, J. R. Schneider, L. 30(4):483–497, 2002. doi:10.1114/1.1467679.
50
Ferrucci, L. Celic, L. M. Taylor, E. Vonesh, G. J. Martin, Tai, N. R., A. Giudiceandrea, H. J. Salacinski, A. M.
and E. Clark. Functional decline in peripheral arterial Seifalian, and G. Hamilton. In vivo femoropopliteal arte-
disease: associations with the ankle brachial index and leg rial wall compliance in subjects with and without lower
symptoms. JAMA 292(4):453–461, 2004. doi:10.1001/jama. limb vascular disease. J. Vasc. Surg. 30(5):936–945, 1999.
292.4.453. doi:10.1016/S0741-5214(99)70020-0.
35 51
Mooney, M. A theory of large elastic deformation. J. Appl. Wensing, P. J., L. Meiss, W. P. Mali, and B. Hillen. Early
Physiol. 11:582–592, 1940. doi:10.1063/1.1712836. atherosclerotic lesions spiraling through the femoral artery.
36
Motomiya, M., and T. Karino. Flow patterns in the human Arterioscler. Thromb. Vasc. Biol. 18(10):1554–1558, 1998.
52
carotid artery bifurcation. Stroke 15(1):50–56, 1984. Wood, N. B., S. Z. Zhao, A. Zambanini, M. Jackson, W.
37
Mozersky, D. J., D. S. Sumner, D. E. Hokanson, and D. E. Gedroyc, S. A. Thom, A. D. Hughes, and X. Y. Xu.
Strandness, Jr. Transeutaneous measurement of the elastic Curvature and tortuosity of the superficial femoral artery: a
properties of the human femoral artery. Circulation possible risk factor for peripheral arterial disease. J. Appl.
46(5):948–955, 1972. Physiol. 101(5):1412–1418, 2006. doi:10.1152/japplphy-
38
Ostchega, Y., R. Paulose-Ram, C. F. Dillon, Q. Gu, and siol.00051.2006.
53
J. P. Hughes. Prevalence of peripheral arterial disease and Wu, S. P., S. Ringgaard, S. Oyre, M. S. Hansen, S. Rasmus,
risk factors in persons aged 60 and older: data from the and E. M. Pedersen. Wall shear rates differ between the
national health and nutrition examination survey 1999– normal carotid, femoral, and brachial arteries: an in vivo
2004. J. Am. Geriatr. Soc. 55(4):583–589, 2007. doi:10.1111/ MRI study. J. Magn. Reson. Imaging 19(2):188–193, 2004.
j.1532-5415.2007.01123.x. doi:10.1002/jmri.10441.
39 54
Papaharilaou, Y., J. A. Ekaterinaris, E. Manousaki, and Younis, H. F., M. R. Kaazempur-Mofrad, R. C. Chan, A. G.
A. N. Katsamouris. A decoupled fluid structure approach Isasi, D. P. Hinton, A. H. Chau, L. A. Kim, and R. D. Kamm.
for estimating wall stress in abdominal aortic aneurysms. Hemodynamics and wall mechanics in human carotid bifur-
J. Biomech. 40(2):367–377, 2007. doi:10.1016/j.jbiomech. cation and its consequences for atherogenesis: investigation of
2005.12.013. inter-individual variation. Biomech. Model. Mechanobiol.
40
Perktold, K., and G. Rappitsch. Computer-simulation of 3:17–32, 2004. doi:10.1007/s10237-004-0046-7.
55
local blood-flow and vessel mechanics in a compliant car- Zeng, D., and R. Ethier. A semi-torsional spring analogy
otid-artery bifurcation model. J. Biomech. 28(7):845–856, model for updating unstructured meshes in 3D moving
1995. doi:10.1016/0021-9290(95)95273-8. domains. Finite. Elem. Anal. Des. 41(11–12):1118–1139,
41
Prendergast, P. J., C. Lally, S. Daly, A. J. Reid, T. C. Lee, 2005.
56
D. Quinn, and F. Dolan. Analysis of prolapse in cardio- Zhang, W., C. Herrera, S. N. Atluri, and G. S. Kassab. Effect
vascular stents: a constitutive equation for vascular tissue of surrounding tissue on vessel fluid and solid mechanics.
and finite element modeling. J. Biomech. Eng. 125(5):692– J. Biomech. Eng. 126(6):760–769, 2004. doi:10.1115/
699, 2003. doi:10.1115/1.1613674. 1.1824128.
42 57
Rhee, K., and S. M. Lee. Effects of radial wall motion and Zhao, S. Z., P. Papathanasopoulou, Q. Long, I. Marshall,
flow waveform on the wall shear rate distribution in the and X. Y. Xu. Comparative study of magnetic resonance
divergent Vascular Graft. Ann. Biomed. 26(6):955–964, imaging and image-based computational fluid dynamics for
1998. doi:10.1114/1.31. quantification of pulsatile flow in a carotid bifurcation
43
Sanctis, J. T. D. Percutaneous interventions for lower phantom. Ann. Biomed. 31(2):962–971, 2003. doi:10.1114/
extremity peripheral vascular disease. Am. Fam. Physician 1.1590664.
58
64(12):1965–1972, 2001. Zhao, S. Z., X. Y. Xu, A. D. Hughes, S. A. Thom, A. V.
44
Smedby, O. Do plaques grow upstream or downstream: an Stanton, B. Ariff, and Q. Long. Blood flow and vessel
angiographic study in the femoral artery. Arterioscler. mechanics in a physiologically realistic model of a human
Thromb. Vasc. Biol. 17(5):912–918, 1997. carotid arterial bifurcation. J. Biomech. 33(8):975–984,
45
Smedby, O., and L. Bergstrand. Tortuosity and athero- 2000. doi:10.1016/S0021-9290(00)00043-9.
sclerosis in the femoral artery: what is cause and what is