Author's Accepted Manuscript

Evaluation of aneurysm-associated wall shearstress related to morphological variations ofcircle of Willis using a microfluidic device

Seong-Won Nam, Samjin Choi, Youjin Cheong,Hun-Kuk Park

PII: S0021-9290(14)00606-XDOI: http://dx.doi.org/10.1016/j.jbiomech.2014.11.018Reference: BM6878

To appear in: Journal of Biomechanics

Accepted date: 12 November 2014

Cite this article as: Seong-Won Nam, Samjin Choi, Youjin Cheong, Hun-KukPark, Evaluation of aneurysm-associated wall shear stress related tomorphological variations of circle of Willis using a microfluidic device, Journalof Biomechanics, http://dx.doi.org/10.1016/j.jbiomech.2014.11.018

This is a PDF file of an unedited manuscript that has been accepted forpublication. As a service to our customers we are providing this early version ofthe manuscript. The manuscript will undergo copyediting, typesetting, andreview of the resulting galley proof before it is published in its final citable form.Please note that during the production process errors may be discovered whichcould affect the content, and all legal disclaimers that apply to the journalpertain.

www.elsevier.com/locate/jbiomech

Evaluation of aneurysm-associated wall shear stress related to morphological variations of circle of Willis using a microfluidic device

Seong-Won Nam*, Samjin Choi*, Youjin Cheong, Hun-Kuk Park

Department of Biomedical Engineering and Healthcare Industry Research Institute, College

of Medicine, Kyung Hee University, Seoul 130-701, Republic of Korea

* The first 2 authors contributed equally to this work.

Address correspondence to:

Hun-Kuk Park, MD, PhD

Department of Biomedical Engineering, College of Medicine, Kyung Hee University, 1

Hoegi-dong, Dongdaemun-gu, Seoul 130-701, Republic of Korea

Tel: +82 2 961 0290

Fax: +82 2 6008 5535

E-mail: sigmoidus@khu.ac.kr

Abstract

Although microfluidic systems have been important tools in analytical chemistry, life

sciences, and medical research, their application was rather limited for drug-screening and

biosensors. Here, we described a microfluidic device consisting of a multilayer micro-

channel system that represented the hemodynamic cerebral vascular system. We analyzed

wall shear stresses related to aneurysm formation in the circle of Willis (CoW) and their

morphological variations using this system. This device was controlled by pneumatic valves,

which occluded various major arteries by closing the associated channels. The hemodynamic

analysis indicated that higher degrees of shear stress occurred in an anterior communicating

artery (ACoA), particularly in the hypoplastic region of the posterior communicating artery

(PCoA) and the P1 segment. Furthermore, occlusion of a common carotid artery (CCA) or a

middle cerebral artery (MCA) increased the shear stress, whereas occlusion of a vertebral

artery (VA) decreased the shear stress. These results indicate that the morphological variation

of the CoW may affect aneurysm formation resulting from increased wall shear stress.

Therefore, the technique described in this paper provides a novel method to investigate the

hemodynamics of complex cerebral vascular systems not accessible from previous clinical

studies.

Keywords: Hemodynamics; Shear stress; Microfluidics; Aneurysm; Circle of Willis

1. Introduction

Atherosclerosis and intracranial aneurysms are common causes of cerebral vascular disease.

Atherosclerosis is artery wall thickening those results from cholesterol and triglyceride

accumulation, while an aneurysm is an abnormal dilation of arteries caused by hemodynamic

stresses. Aneurysms are estimated to occur in 6% of the general population, and unruptured

aneurysms are normally asymptomatic (Rinkel et al., 1998). However, subarachnoid

hemorrhage from a ruptured cerebral aneurysm results greater than 50% mortality and is

more frequent than other cerebral diseases such as brain tumors or multiple sclerosis (Hop et

al., 1997; Schievink et al., 1997). A subarachnoid hemorrhage is often the initial clinical

presentation of asymptomatic unruptured aneurysms. The typical symptoms of subarachnoid

hemorrhage include severe headache, loss of consciousness, nausea, vomiting and

meningismus without infection. Clinical findings from an international study of unruptured

intracranial aneurysms (ISUIA) in 1998 indicated that aneurysms smaller than 10 mm in

diameter had an average 0.05% annual rupture rate, whereas patients with a history of

subarachnoid hemorrhage had a rupture rate of 0.5% (ISUIA, 1998). The annual rupture rate

was about 1% for aneurysms larger than 10 mm in diameter (ISUIA, 1998). Two invasive

neurosurgical methods are used to treat intracranial aneurysms: endovascular coiling and

neurosurgical clipping. A retrospective study reported that patients that underwent

endovascular coiling had a higher survival rate than those that were treated with

neurosurgical clipping, although the risk of bleeding was higher after endovascular coiling

(Molyneux et al., 2005).

Aneurysm formation has been linked to both morphological variations of circle of Willis

(CoW) and gender. Aneurysms in the anterior communicating artery (ACoA) or the internal

carotid artery (ICA)- posterior communicating artery (PCoA) were identified more frequently

in the presence of an abnormal or an absent A1 segment or fetal posterior cerebral artery

(PCA). An absent A1 segment was more common in men than women while fetal PCA

variations occur more frequently in women (Horikoshi et al., 2002). The correlation between

high incidence of aneurysms and morphological CoW variations may have been caused by

disruptions in blood flow due to anatomical changes in the CoW. Indeed, variations in the

fetal PCA increased flow rates in the ipsilateral ICA, whereas absence of the A1 segment

induced lower flow rates in the ipsilateral ICA (Hendrikse et al., 2005). In comparison,

saccular aneurysms developed at vessel bifurcations, where shear stress against the arterial

wall was greatest. It has been proposed that the frictional forces resulting from high wall

shear stress and hypertension may have triggered the formation of aneurysms in the arterial

wall (Inci et al., 2000; Gao et al., 2008). Nevertheless, the direct effect wall shear stress

resulting from specific CoW morphological variations has on aneurysm formation and has

never been elucidated.

In this study, we present a novel technique that uses a microfluidic system to study the

relationship between wall shear stress and aneurysm formation. A cerebral vascular system

that included the CoW was constructed, and pneumatic-controlled microvalves were designed

to control microchannels within the system (Nam et al., 2007, 2009), which could close and

occlude the arteries. Furthermore, the system modeled three representative morphological

variants and their effect on wall shear stress. To our knowledge, wall shear stress has never

been studied using a microfluidic system, and the relationship between wall shear stress and

CoW morphology at various carotid artery occlusion sites has also never been studied. Thus,

this study was designed to better understand the pathogenesis and functional anatomy of

hemodynamic-correlated aneurysm formation in the cerebral vascular system.

2. Experimental methods

2.1. Microfluidic device and microvalve construction

The microfluidic device was constructed using soft lithography, consisting of two layers of

PDMS (Dow-Corning, Cortland, NY), as described previously (figure 1) (Nam et al., 2007,

2009). The thick top layer and thin bottom layer were prepared with a curing agent-PDMS

ratio of 1:10 and 1:30, respectively. The thick layer was aligned with the thin layer, and the

two layers were in contact. The PDMS was joined with a microscope slide, and an

irreversible bond was formed using O2 plasma cleaner (Femto Science, Hwaseong, Korea).

Pneumatic valves were connected to solenoid valves (LHDA0523112H; Lee Company,

Westbrook, CT), which were used to pressurize or close the PDMS valves. Solenoid valves

were controlled with a one-chip, 8-bit microcontroller (PIC16F873A; Microchip Technology

Inc., Chandler, AZ) (Choi et al., 2006). The N2 gas pressure was maintained at 30 psi by a

mini-pressure regulator.

2.2. Microfluidic system setup and operation

The cerebral artery dimensions were based on the average of literature data (Alastruey et

al., 2007). The inlet flow was generated using two syringe pumps (Kd Scientific, Holliston,

MA) connected to two common carotid artery (CCA) channels (52.97 �L/min) and two

vertebral artery (VA)-channels (16.77 �L/min) through plastic tubing (Tygon; Saint-Gobain

Co., Courbevoie, France). The channels were filled with a synthetic blood solution,

containing 40% glycerol and orgasol particles (�=3.5–6.5 �m; Arkema, Colombes, France),

which resembled the density (�=1053 kg/m3) and viscosity (�=3 mPa) of human blood

(Chung et al., 2010). Flow rates were measured using the volume of fluid from the outlets

and confirmed by a Laser-Doppler Flowmeter (LDF; Transonic Systems Inc., Ithaca, NY).

2.3. Computational hemodynamic remodeling

The three-dimensional model was constructed with SolidWorks 2010 software

(SolidWorks Corp., Waltham, MA) according to the geometric features of the microfluidic

cerebral vascular system and then transformed to finite element analysis software (COMSOL

Multiphysics 4.1; COMSOL Inc., Burlington, MA). One normal CoW model and three CoW

variant models were constructed. A structured mesh with approximately 3×106 triangular

elements was generated to span the entire fluid model. The no-slip velocity condition at solid

boundaries was applied to all the microchannel walls. All meshes were assigned to vascular

walls or microchannel inlets and outlets. The blood flow was modeled as a steady laminar

and incompressible Newtonian fluid. The governing equations were the continuity equations

(mass, Eq (1)) and the Navier-Stokes equations (momentum balance, Eq (2));

0i

i

ux

��

� (1)

2

2

1i ij

j i j

u upux x x

��

� �� ��� � �� �� � � �

(2)

where ui is the velocity component in the i direction, p is the pressure, � is the fluid density,

and μ is the kinematic viscosity. The fluid properties were evaluated at an operating

temperature of 293K. The flow field was modeled as steady and laminar using Eqs. (1) and

(2), and the inlet/outlet boundary conditions were set as mass inflow and outflow. The

synthetic blood solution (�=1037 kg/m3; �=4.1 mPa) modeled blood in the system. The

Reynolds number was 8 for CCA and 20 for VA. Numerical simulations were performed

with COMSOL Multiphysics. Velocities identical to those used for the experimental setup of

the microfluidic system were applied to afferent vessels. The efferent vessel velocities were

set to match the experimental results from each of the conditions. Artery occlusion was

implemented by adjusting the mesh wall settings. Iterative computation was conducted in 500

equal time steps. The velocity, pressure and shear stress of the microfluidic cerebral vascular

system was modeled using computational hemodynamic analysis.

3. Results and discussion

To elucidate the correlation between high wall shear stress and aneurysm formation, in this

study, the relationship between morphological CoW variation and shear stress in efferent

arteries was investigated. The shear stress resulting from main artery occlusion was also

analyzed and compared with morphological variations. This study used a novel microfluidic

system model to determine flow dynamic data to create numerical simulations for the

analysis of wall shear stress.

Figure 1 shows the representative magnetic resonance angiography image of the CoW and

the microfluidic chip, where the cerebral blood flow was analyzed. The dimensions of the

microfluidic channels were 1/10 of the dimensions derived from clinical data of the

corresponding vessels (Alastruey et al., 2007). The applied inlet flow rates were 52.97

�L/min for the CCAs and 16.77 �L/min for the VAs, to account for the reduced channel

dimensions. The flow rates of 4 efferent artery pairs were measured by analyzing fluids

collected directly from the outlets and confirmed using the Laser-Doppler Flowmeter (LDF).

The flow rates measured by outflow volume were 90.8% accurate (Pearson correlation;

r=0.9076) and those measured by LDF were 98.9% accurate (r=0.9898) when compared to

the clinical average (Chung et al., 2010). Another characteristic of this model system was that

the channels could be closed using the pneumatic micro-valves to mimic artery occlusion.

Figure 2 shows the microfluidic system design, microfluidic chip, and three-dimensional

models for the analysis of computational fluid dynamics (CFD). Only 42% of the population

has a complete CoW structure where underdeveloped or absent blood vessels have been

identified (Eftekhar et al., 2006; Papantchev et al., 2007). Among them, 9% have variations

in the ACoA and 15-22% of individuals have variations in the P1 segment or PCoA (Eftekhar

et al., 2006; Papantchev et al., 2007). Thus, these three representative variations were

modelled to determine the correlation between morphological variations and wall shear stress

(Fig. 2).

Figure 3 shows an example of hemodynamic analysis based on flow rate measurement in

the microfluidic system. The flow velocity, static pressure, and shear stress were analyzed by

finite element analysis software (COMSOL Multiphysics) based on the flow rate information

measured in the microfluidic cerebral vascular system (Fig. 1).

Figure 4 shows the representative contours of flow velocity in the CoW. When the CCA

was occluded, the flow rate of the ACoA and A1 segments increased in the complete CoW

and in the three morphological variations. However, the greatest increase in the CoW flow

rate occurred in the model with an absent P1 segment (Fig. 4H).

Figure 5 shows the wall shear stress in the complete CoW and the morphological variants.

In the complete CoW without occlusion, the shear stress was bilaterally symmetrical (Fig. 5).

The bifurcations between the ACoA and ACA, CCA and MCA, and VA and P1 segments

showed the highest shear stress, whereas the centers of the ACA, MCA, PCoA, and PCA

showed low shear stress. The average shear stress value in the microfluidic system was

0.39�0.056 Pa (3.89 dyn/cm2), which was comparable to published experimental values, in

which shear stress on arterial walls was estimated up to 16 dynes/cm2 (Papaioannou et al.,

2005). The duplicated ACoA did not affect the shear stress when compared to the normal

CoW, in which the maximum shear stress was 0.48�0.057 Pa (Fig. 5). In comparison, an

absent PCoA or P1 segment increased the shear stress by approximately 18% or 59.1%,

respectively, when compared to the complete CoW (Fig. 5). Shear stress on the ACoA, PCoA,

and P1 segments increased to 1.57�0.15 Pa in the CoW morphological variant without the P1

segment (Fig. 5), which may have increased the probability of aneurysm formation at those

positions.

Figure 6 shows shear stress distributions as a function of pulsatile input. Shear stress

distributions varied with the applied pressure, but were not significantly different than those

under constant pressure in both the complete CoW and the CoW variants. Shear stress

distributions at maximum pressure (t=0.65) were comparable to distributions at constant

pressure (Fig. 5).

The physical parameters of velocity (m/s) and atrial pressure (Pa) were applied to computer

simulations to obtain shear stress, and the data were analyzed to evaluate the relationship

among blood pressure, wall shear stress, and aneurysm formation in cases of morphological

CoW variations in addition to major artery occlusion (Figs. 5 and 6). This study determined

the predicted location and degree of risk for aneurysm formation based on the flow

mechanics of blood. Until now, the relationship between morphometric variations in the CoW

and aneurysm formation has never been quantitatively analyzed, and some of the results in

this study were verified with clinical findings. Thus, changes in shear stress due to occlusion

of the CCA, the VA, or efferent arteries can be considered important enough to deserve

further studies with clinical validation. The purpose of this analysis was to elucidate the

relationship between wall shear stress and aneurysm formation. The research answered the

following questions: 1) Does high wall shear stress cause aneurysms? 2) Are morphological

CoW variations related to aneurysm formation? 3) Is artery occlusion related to aneurysm

formation?

Figure 5 shows that higher shear stress occurred in the ACoA, at the bifurcation of the MCA

to the CCA, and at VA branching points, which are consistent with aneurysm formation

positions (Rinkel et al., 1998). Therefore, it can be concluded that high wall shear stress may

correlate with aneurysm formation. Also, PCoA hypoplasticity and the presence of a P1

segment increased the shear stress, which may also increase the possibility of aneurysm

formation. This figure also showed that CCA or MCA occlusion increased shear stress,

whereas VA occlusion decreased shear stress. This result implies that aneurysm formation is

dependent on artery occlusion. Artery occlusion generally increased the shear stress,

especially in cases of P1 segment hypoplasticity (Fig. 5). These results indicate that

morphological CoW variations can increase the risk of cerebral ischemia as well as

aneurysms that are caused by high wall shear stress. Therefore, CoW morphological

variations should be considered in future clinical studies that seek to determine risk factors

for aneurysm formation.

Although the microfluidic device provides the pressure and flow data for the detailed three-

dimensional CFD modeling, there are some limitations such as minimized and simplified

replications of the cerebral vasculature, neglecting the property of vessel walls to realize the

CBF, neglecting the complicated feature of flow including the reflected pressure by the

channel shapes, implementation of only major arteries in the CoW system, and no endothelial

cells in blood vessel. However, despite these limitations the flow rate in the micro-channels

achieved 98.9% accuracy compared to the clinical data (Chung et al. 2010). Also, the

microfluidic model system showed that the flow in the microchannels had similarity to non-

Newtonian fluid by using blood mimicking solution with orgasol particles (Chung et al.

2010). Therefore the microfluidic device would be the superior pre-screening system to

reduce the gap between the computed modeling and clinical reality (Hassan et al., 2004). In

this study, a microfluidic system was used to study the hemodynamics of aneurysm

formation. Since the computer simulation depended on the boundary conditions, the

microfluidic system could be used to confirm simulated results. Despite of the square channel

and irregular cross-section, the relatively cross-sectional area of the channels was considered

and the influence of flow rate on arterial pressure was comparable with clinical data.

Therefore, the model system had strong agreement between the computed results and

clinically determined values.

5. Conclusion

In this paper, we describe a microfluidic system for the investigation of wall shear stress in

the cerebral vascular system. The plasticity of the microfluidic system and integration of

pneumatic-controlled microvalves allowed for an analysis of the relationship between shear

stress in morphological variation of CoW structure and occlusion of major arteries. Higher

wall shear stress was found in the ACoA of the complete CoW and the CoW without PCoA

and P1 segments, which is consistent with clinical aneurysm formation. Occlusion of the

CCA and the MCA increased wall shear stress in the CoW, while occlusion of the VA

decreased the shear stress. The analysis of wall shear stress in this study demonstrated a

correlation between high wall shear stress and aneurysm formation.

Although this study verified that the microfluidic system is a valid model for the study of

cerebral hemodynamics, the relation between vascular endothelium and aneurysm formation

was not taken into account. Therefore, the next step would include using a microfluidic

device that accounts for the effect of wall shear stress on endothelial cells. This would lead to

an even better understanding of the wall shear stress in aneurysm formation.

Acknowledgement

This study was supported by a grant from the Korean Health Technology Research &

Development Project, by the Ministry of Health & Welfare, Republic of Korea (A110216).

Competing Interests: The authors declare that no competing interests exist.

References

Alastruey, J., Parker, K.H., Peiro, J., Byrd, S.M., Sherwin, S.J., 2007. Modeling the circle of

Willis to assess the effects of anatomical variations and occlusion on cerebral flow.

Journal of Biomechanics 40, 1794-1805.

Choi, S., Jiang, Z., 2006. A novel wearable sensor device with conductive fabric and PVDF

film for monitoring cardiorespiratory signals. Sens. Actuators A: Phys. 128, 317-326.

Chung, E.M.L., Hague, J.P., Chanrion, M.A., Ramnarine, K.V., Katsogridakis, E., Evans,

D.H., 2010. Embolus trajectory through a physical replica of the major cerebral arteries.

Stroke 41, 647-652.

Eftekhar, B., Dadmehr, M., Ansari, S., Ghodsi, M., Nazparvar, B., Ketabchi, E., 2006. Are

the distributions of variations of circle of Willis different in different populations?-Results

of an anatomical study and review of literature. BMC Neurology 6, 22.

Gao, L., Hoi, Y., Swartz, D.D., Kolega, J., Siddiqui, A., Meng, H., 2008. Nascent aneurysm

formation at the basilar terminus induced by hemodynamics. Stroke 39. 2085-2090.

Hassan, T., Timofeev, E.V., Saito, T., Shimizu, H., Ezura, M., Tominaga, T., Takahashi, A.,

Takayama, K., 2004. Computational replicas: Anatomic reconstructions of cerebral

vessels as volume numerical grids at three-dimensional angiography. Am. J. Neuroradiol.

25, 1356-1365.

Hendrikse, J., van Raamt, A.F., van der Graaf, Y., Mali, W.P., van der Grond, J., 2005.

Distribution of cerebral blood flow in the circle of Willis. Radiology 235, 184-189.

Hop, J.W., Rinkel, G.J., Algra, A., van Gijn, J., 1997. Case-fatality rates and functional

outcome after subarachnoid hemorrhage: a systematic review. Stroke 28, 660-664.

Horikoshi, T., Akiyama, I., Yamagata, Z., Nukui, H., 2002. Retrospective analysis of the

prevalence of asymptomatic cerebral aneurysm in 4518 patients undergoing magnetic

resonance angiography-when does cerebral aneurysm develop? Neurol. Med. Chir. 42,

105-112.

Inci, S., Spetzler, R.F., 2000. Intracranial aneurysms and arterial hypertension: a review and

hypothesis. Surg. Neurol. 53, 530-540.

International Study of Unruptured Intracranial Aneurysms Investigators (ISUIA), 1998.

Unruptured intracranial aneurysms-risk of rupture and risks of surgical intervention. N.

Engl. J. Med. 339, 1725-1733.

Molyneux, A.J., Kerr, R.S.C., Yu, L.M., Clarke, M., Sneade, M., Yarnold, J.A., Sandercock,

P., 2005. International subarachnoid aneurysm trial (ISAT) of neurosurgical clipping

versus endovascular coiling in 2143 patients with ruptured intracranial aneurysms: a

randomized comparison of effects on survival, dependency, seizures, rebleeding,

subgroups, and aneurysm occlusion, Lancet 366, 809-817.

Nam, S.W., Kim, S.T., Lee, K.M., Kim, S.H., Kou, S., Lim, J., Hwang, H., Joo, M.K., Jeong,

B., Yoo, S.H., Park, S., 2009. N-methyl-D-aspartate receptor-mediated chemotaxis and

Ca2+ signaling in Tetrahymena pyriformis. Protist 160, 331-342.

Nam, S.W., van Noort, D., Yang, Y., Park, S., 2007. A biological sensor platform using a

pneumatic-valve controlled microfluidic device containing Tetrahymena pyriformis. Lab.

Chip 7, 638-640.

Papaioannou, T.G., Stefanadis, C., 2005. Vascular wall shear stress: Basic principles and

methods. Hellenic J. Cardiol. 46, 9-15.

Papantchev, V., Hristov, S., Todorova, D., Naydenov, E., Paloff, A., Nikolov, D., Tschirkov,

A., Ovtscharoff, W., 2007. Some variations of the circle of Willis, important for cerebral

protection in aortic surgery - a study in Eastern Europeans. Eur. J. Cardiothorac. Surg. 31,

982-989.

Rinkel, G.J., Djibuti, M., Algra, A., van Gijn, J., 1998. Prevalence and risk of rupture of

intracranial aneurysms: a systematic review. Stroke 29, 251-256.

Schievink, W.I., 1997. Intercranial aneurysm. N. Engl. J. Med. 336, 28-40.

Figures

Fig. 1. Magnetic resonance angiography (MRA) image, CAD design of proposed model and microfluidic chip of the cerebral vascular circle of Willis (CoW) system. Occlusion of arteries was induced by pneumatic valves, shown as red lines on the CAD design. ACA: anterior cerebral artery, A1, A2: segments of ACA, ACoA: anterior communicating artery, BA: basilar artery, CCA: common carotid artery, ECA: external carotid artery, ICA: internal carotid artery, MCA: middle cerebral artery, PCA: Posterior cerebral artery, P1: segment of PCA, PCoA: posterior communicating artery, VA: vertebral artery.

Fig. 2. Design, microfluidic chip and CFD models of the CoW and morphological variations; (A) complete CoW, (B) CoW with duplicated ACoA, (C) CoW without the PCoA, and (D) CoW without the P1 segment. E: Elements.

Fig. 3. Hemodynamic analysis based on flow rate measurements in the microfluidic system. (A) Microfluidic model of CoW. (B) After image processing. (C) COMSOL segmentation. (D) Flow velocity field. (E) Static pressure distribution. (F) Wall shear stress distribution in a complete CoW.

Fig. 4. Representative images of velocity contours in the CoW and the variants with and without occlusion. (A-D) Complete CoW (A) without occlusion, (B) with LCCA occlusion, (C) with VA occlusion, (D) with LMCA occlusion. (E-F) Velocity contours with RCCA occlusion (E) in a complete CoW, (F) in a CoW with a duplicated ACoA, (G) in a CoW with absent PCoA, (H) in a CoW with absent P1 segment. Scale bar = 1 mm.

Fig. 5. Shear stress with artery occlusion in a complete CoW and morphological variants. First row: complete CoW, second row: CoW with a duplicated ACoA, third row: CoW with no PCoA, fourth row: CoW with absent P1 segment. First column: without occlusion, second column: with right CCA occlusion, with VA occlusion, with left MCA occlusion.

Fig. 6. Shear stress distributions and pressure waveforms for a complete CoW and a CoW with absent P1 segment and MCA occlusion. Shear stress distributions for a complete CoW model with respect to maximum pressure, t = 0.65 (A) and minimum pressure, t = 1.0 (B). (C) Input pressure waveform at middle ACoA point. Shear stress distributions for a left MCA occlusive CoW model with respect to maximum pressure, t = 0.65 (D) and minimum pressure, t = 1.0 (E). (F) Shear stress distributions at middle ACoA point.