In vitro study of hemodynamic treatment improvement: Hunterian ligation of a fenestrated basilar artery aneurysm after coiling


Hunterian ligation affecting hemodynamics in vessels was proposed to avoid rebleeding in a case of a fenestrated basilar artery aneurysm after incomplete coil occlusion. We studied the hemodynamics in vitro to predict the hemodynamic changes near the aneurysm remnant caused by Hunterian ligation. A transparent model was fabricated based on three-dimensional rotational angiography imaging. Arteries were segmented and reconstructed. Pulsatile flow in the artery segments near the partially occluded (coiled) aneurysm was investigated by means of particle image velocimetry. The hemodynamic situation was investigated before and after Hunterian ligation of either the left or the right vertebral artery (LVA/RVA). Since post-ligation flow rate in the basilar artery was unknown, reduced and retained flow rates were simulated for both ligation options. Flow in the RVA and in the corresponding fenestra vessel is characterized by a vortex at the vertebrobasilar junction, whereas the LVA exhibits undisturbed laminar flow. Both options (RVA or LVA ligation) cause a significant flow reduction near the aneurysm remnant with a retained flow rate. The impact of RVA ligation is, however, significantly higher. This in vitro case study shows that flow reduction near the aneurysm remnant can be achieved by Hunterian ligation and that this effect depends largely on the selection of the ligated vessel. Thus the ability of the proposed in vitro pipe-line to improve hemodynamic impact of the proposed therapy was successfully proved.

Int J Artif Organs 2014; 37(4): 325 - 335




Leonid Goubergrits, Andreas Spuler, Jens Schaller, Nils Wiegmann, Andre Berthe, Hans-Christian Hege, Klaus Affeld, Ulrich Kertzscher

Article History


Financial Support: This study was supported by the German Research Foundation (DFG project KE 900/9-1).
Conflict of Interest: None.

This article is available as full text PDF.

Download any of the following attachments:


Cerebral aneurysm rupture causing life-threatening subarachnoid hemorrhage (SAH) is associated with high mortality and morbidity (1). The major therapy options are surgical clip placement or endovascular coil embolization, both aiming at separating the aneurysm from the parent vessel blood flow and thereby preventing rebleeding (2). Recently, a new endovascular aneurysm treatment using a flow diverting stent has been introduced (3). The idea behind this method is to avoid or reduce blood inflow from the parent vessel into the aneurysm. Further treatment options include the use of aneurysm filling with hardening polymers or glue (2). Hunterian ligation, which is considered the father of modern neurosurgical clipping, is also an option (4, 5). Choosing the optimal treatment depends on many risk factors associated with SAH, including sex, age, ethnicity, aneurysm size, location, shape, and patient history (2, 6). Due to the huge inter-patient variability, the choice of therapy option is best left to a team of clinicians familiar with neurovascular surgery, endovascular treatment techniques, and neurologic critical care (7). The objective of treatment selection is to predict patient-specific post-treatment changes of intra-aneurysmal and parent vessel hemodynamics due to luminal changes caused by artificial devices (coils, stents, flow diverting stents, or clips). This requires an image-based flow analysis using an experimental model or computational fluid dynamics. Here, the expertise of engineers in hemodynamics – of biofluid mechanics, for example – comes into play. Without a hemodynamic analysis before the intervention, the effect of such an treatment decision, even in the case of Hunterian ligation, can be clinically unexpected and disastrous (8).

In this study, we describe an experimental pipeline using an in vitro model to investigate the patient-specific hemodynamic impact of Hunterian ligation of a vertebral artery (VA) on a fenestrated basilar artery after incomplete coiling of an aneurysm on one of the fenestra channels. We simulated Hunterian ligation of either the left or the right VA. The aim of Hunterian ligation is flow reduction near the aneurysm remnant. In this study, the impact of Hunterian ligation on hemodynamics near the aneurysm remnant at a fenestrated basilar artery was investigated for the first time. The analysis of the hemodynamics was done by means of particle image velocimetry (PIV).


Data and reconstruction, model cast

An aneurysm was detected on a fenestrated basilar artery of a 45-year-old female. Computed tomography (CT) angiography was performed by means of a Siemens CT VA0 device (Siemens, Forchheim, Germany) using an 18 s 1.25 H21f sequence with 512 x 512 px in-plane slice resolution and 1.25 mm slice thickness, resulting in the voxel size of 0.5 mm x 0.5 mm x 0.9 mm. Due to the low resolution of the CT data and the difficult segmentation caused by the proximity of the skull base, these data were not accurate enough to allow for an appropriate three-dimensional (3D) reconstruction. Figure 1a shows rough CT-based reconstruction giving a 3D representation of the untreated aneurysm. At the neuroradiological department of the Helios Hospital, Berlin-Buch, Germany, the aneurysm was completely occluded by coiling (Fig. 1b).

a) CT slice with a reconstructed aneurysm at fenestrated basilar artery. b) Angiogram of the coiled aneurysm after a successful exclusion of the aneurysm from blood circulation. c) Angiogram showing failed (contrast agent enters the aneurysm) coiling therapy 21 months after the treatment.

However, 21 months later, routine follow-up 3D rotational digital subtraction angiography (3DRA with 126 frames, 1.6° between two frames) found incomplete aneurysm occlusion (Fig. 1c). The 3DRA high-resolution data set was preliminary reconstructed on a Leonardo InSpace 3D workstation (Siemens, Forchheim, Germany) with an isotropic voxel size of 0.22 mm. It allowed for 3D vessel segmentation and reconstruction using the program ZIB-Amira (Zuse-Institute Berlin, Germany): thresholding and semi-automatic image segmentation were performed using voxel labeling with sub-voxel accuracy. The surface was generated from these labeled data (Fig. 2a) according to a procedure described in our earlier work (9). The inlet diameter of the left vertebral artery (LVA) was DLVA = 3.12 mm, whereas the inlet diameter of the right vertebral artery (RVA) was DRVA = 2.62 mm.

a) Reconstructed post-coil surface with aneurysm remnants. Black arrows show flow direction. b) and c) Dashed lines show locations of two laser light sheets (A-plane and B-plane) acquired by PIV.

During reconstruction we preserved inlet segments as long as possible (8 x DLVA and 7.6 x DRVA) and reconstructed the aneurysm remnant. A two-fold scaled-up transparent thin-walled (1.5 mm thickness) model allowing in vitro clipping (vessel ligation) was manufactured from silicon rubber (Elastosil 601 RT, refraction index of n = 1.4, Wacker Chemie, Munich, Germany) with a precision of 0.1 mm according to a procedure previously described (10). The silicon model (Fig. 3) has a luminal volume of 7.2 mL. The resulting volume distensibility (E = dp*V/dV, where V is the luminal volume, and dV is the volume change due to the pressure change dp) was E = 110 kPa, which correlates well with the literature data (11). This means a change of vessel radius by 2.5% due to a pressure increase of 40 mmHg (5332 Pa). Volume distensibility was measured by measuring the pressure (Multi-Handy 3050 manometer, Hydrotechnik, Limberg, Germany) in the clamped model due to insertion of the fluid into the model by a syringe.

a) Experimental setup including (1) PC for data storing, (2) computer controlled piston pump, (3) throttles to set flow rates, (4) rigid inlet ducts, (5) collecting reservoirs, (6) laser, (7) optic generating laser light sheet, (8) flow rate probe Transonic, (9) camera, and (10) Plexiglas® box containing silicone model and filled with glycerol/water mixture. Black arrows show flow directions. b) In vitro scaled inlet flow waveform. Four time steps (end-diastole – t1 (176 ml/min); acceleration phase – t2 (266 ml/min); peak-systole – t3 (388 ml/min); deceleration phase – t4 (266 ml/min)) represent characteristic flow states during heart cycle.

Experiment setup

Figure 3a depicts the experimental setup. The vessel model was put into a transparent box with long rigid straight ducts (L = 140 mm) mounted at inlets and filled with a test fluid also serving as blood-analog. Long inlets ensured fully developed flow at the inlets of the vessel model. The test fluid was a glycerin/distilled water mixture (37.5% mass fraction of the 99.5% pure glycerin) with a refractive of n = 1.4 and density of 1100 kg/m3. A dynamic viscosity of the test fluid of 3.47 mPa·s at room temperature of 24°C was ascertained by an Ubbelohde viscosimeter 50101 (Schott, Mainz, Germany) with a 1% accuracy. Unsteady flow was generated by a computer-controlled piston pump that was built in-house, described earlier (12). Outlets were connected to a measuring cylinder. The model includes three outlets: the major outlet is the outlet of the basilar artery, which has also two side branches (left and right anterior inferior cerebellar arteries). The flow distribution between outlets was kept constant according Murray’s law (13). Since outlets were far away from the region of interest (~10·diameters of the basilar artery = 80 mm) and the flow rate of both side branches together was only 5% of the total flow rate for both VA, outlets do not affect investigated flow conditions.

Patient-specific flow data were not available. Therefore, a total mean flow rate of 133 ml/min derived from magnetic resonance imaging (MRI) for both VAs (flow in basilaris artery) was simulated for an original unclamped situation (14, 15). The waveform (Fig. 3b) was taken from phase contrast MRI measurements of patient number four in the publication of Rayz et al (16). Flow parameters in the in vitro model were set taking into account the two-fold model scale, the Reynolds-number similarity (Re = U·D/ν = constant, where U is mean inlet velocity, D is inlet diameter, and ν is kinematic viscosity), and the Strouhal-number similarity (St = f·L/U = constant, where f is the heart rate, L is a characteristic length - L = D). Thus, in vitro flow rate was set to 266 ml/min and the period to 4 s. Flow rate was controlled by means of the Transonic T106 flow meter (Transonic Systems, Ithaca, NY, USA) using the in-line flow probe 4N34 with 7% accuracy. Flow rate ratios were adjusted with throttles considering Murray’s law (13). According to the diameters of LVA and RVA, flow rates in LVA/RVA were 63.2%/36.8%. Accuracy of flow rate setting was ±3 ml/min. Reynolds- and Strouhal-numbers allows to calculate also a Womersley number defined as α = (2·π·Re·St)1/2, which was equal 1.7 in RVA and 2.1 in LVA. The cycle averaged Reynolds numbers for LVA and RVA inlets varied between 100 and 420.

Three setups were investigated: the original unclamped model, the model with ligated LVA, and the model with ligated RVA. Since the flow rate after Hunterian ligation can be reduced due to the treatment, two mean post-ligation flow rates were simulated for both clamped configurations:

the flow rate of the respective unclamped LVA (in vitro 168 ml/min) or RVA (in vitro 98 ml/min) meaning a reduction of the mean total flow rate in the basilaris artery due to clamped vessel and;

a total mean flow rate of both VA simulated in the unclamped configuration (in vitro 266 ml/min) meaning an increase of the flow rate in the unclamped vessel in order to compensate the flow in the clamped vessel.

PIV measurements

Standard 2D-PIV (17) was used to investigate the flow in the fenestra vessel segment with aneurysm neck remnant (Fig. 2a). The fenestra vessel segment (FVS) corresponding to the left VA is called LFVS, whereas the opposite fenestra vessel segment corresponding to the right VA and showing an irregular shape is called RFVS (Fig. 2a). Two orthogonal planes (A and B) were investigated (Fig. 2). Tracer particles with diameters of about 14 µm and a density of 1700 kg/m3 (SH400S33; Potters Industries, Malvern, PA, USA) were added to the test fluid. Relaxation time describing the ability of particles to follow the flow was very good with tR = 2*10-6 s (18). A Finesse 4 W laser with a wave length of 532 nm (Laser Quantum, Stockport, UK) generated a light sheet with a thickness of 1 mm. Movements of the illuminated particles were recorded with 1000 fps using a high speed camera Motion Pro X3 (Integrated Design Tools Inc., Tallahassee, FL, USA) with a 50 mm, 1:1.8 NIKKOR lens (Nikon, Tokyo, Japan) triggered by the pump. Images were acquired with the resolution of 38 px/mm. Vector fields were calculated from image pairs using a cross-correlation multi-pass technique (64 x 64 px followed by a window of 32 x 32 px with 50% overlap) implemented in the software Davis 7.2 (LaVision, Göttingen, Germany). Sequences of 50 vector fields were time averaged resulting in 80 image sequences for a heart cycle (4 s in vitro).

As mentioned above, five experimental flow setups were investigated. Inter-experimental reproducibility was proved. Figure 4 shows in vivo velocity profiles with root-mean-square error bars in three selected positions. These error values (mean error of 0.01 m/s) showing very good reproducibility are representative for different flow configurations and time steps. The error is relatively low, considering the range of acquired velocity magnitudes (Vmax = 0.2 m/s), and correlates well with a theoretical PIV accuracy of 0.1 px (0.006 m/s for our settings) (19). Consequently, we did not reproduce error bars in further figures. Additionally max-min-images visualizing particle path lines were generated. For each image sequence, maximal and minimal gray values for each pixel were determined. The difference between them generates a max-min-image representatively shown in the case of the unclipped situation. The combination of PIV and max-min images based on two approaches better represents the flow field. Only four representative time steps (Fig. 3b) are shown in the results.

Flow profiles with error bars in the A-plane for time step t2 with in vivo flow rate of 133 ml/min at three specific positions during original unclamped flow configuration. Dashed lines mark vessel segment centers.


Velocity vector fields are shown in Figures 5 (A-plane) and 7 (B-plane). Additionally, Figure 6 underlines quantitative differences between different flow conditions shown in Figure 5 by a direct comparison of selected flow profiles. Figures 5a and b show the flow in the original situation: the flow is characterized by a vortex at the RVA junction site caused by cross-section enlargement at this point. The vortex (recirculation zone) exists during the whole heart cycle and reduces the effective cross section for flow into the corresponding vessel segment RFVS. The vortex shape varies slightly with flow rate due to heart cycle (Fig. 5b).

Flow fields in A-plane visualized at four characteristic time steps t1-t4 with a total in vivo flow rate of 133 ml/min in all three configurations: a) and b) original unclamped, c) ligated LVA, and d) ligated RVA. The aneurysm remnant is beneath the plane. Grey areas mark the recirculation zone (vortex).

Flow profiles at four characteristic time steps t1-t4 from up to down with the total in vivo flow rate of 133 ml/min at three specific positions: a) original unclamped flow configuration, b) superimposed profiles of three configuration (original unclamped – dashed; ligated LVA – black; ligated RVA – gray line), and c) flow profiles of the ligated LVA configuration with two flow rates in the RVA (in vivo 133 ml/min – black line and in vivo 49 ml/min – gray line). Velocity profiles of the lower flow rate were scaled by a factor of 133/49 for a direct comparison.

In contrast, LVA flow stream enters LFVS (Figs. 5a and b) without any disturbances. This is because LVA is better aligned to the corresponding fenestra vessel section LFVS (Fig. 1c). A minor fraction of the flow stream is strongly deflected and produces higher velocity magnitudes near the inner fenestra wall (Fig. 6a). During the heart cycle major flow features remain the same. Velocity profiles, however, show nearly parabolic profiles with a single peak during diastolic flow and accelerating phase (t1 and t2), whereas profiles during peak systole flow and decelerating phase (t3 and t4) are more irregular with two peaks (Fig. 6a).

The recirculation zone increases in both configurations of unilateral clipping (Figs. 5c and d). This is more easily seen at peak systole (time step t3 in Figs. 5b, c, and d). The effect is greater in the case of RVA clipping (Fig. 5d) due to flow stagnation in the RVA. The vector field displays more homogeneous velocity magnitudes for the unclipped configuration compared to the vector fields after unilateral clipping: due to the retained total flow rate, velocity magnitudes in the unclamped inlet increase, whereas flow in the clamped vessel stagnates. Because of the misalignment of the RVA with respect to the right section of the fenestra (RFVS), the flow jet impinges on the fenestra wall after the LVA ligation. At the impingement point the flow is divided. In the left section of the fenestra (LFVS), flow generates a jet crossing from the inner to the outer wall of the segment. This generates disturbed flow compared to the original unclipped configuration. The flow disturbance is more pronounced in plane B (Fig. 7c). In contrast, RVA ligation causes flow in LFVS, which is very similar to the flow in the unclamped situation.

Flow field in the B-plane at four characteristic time steps t1-t4 with the total in vivo flow rate of 133 ml/min in all three configurations a) and b) original unclamped, c) ligated LVA, and d) ligated RVA. Grey frame (see t1, b) shows the region used to quantify the flow reduction due to ligation near the aneurysm remnant.

The flow profiles at different positions show the quantitative impact of the treatment options (Fig. 6c). Flow stagnates in the region of the ligated VA inlets. A shift of the velocity peak to the outer wall can be identified clearly. In the case of the ligated RVA flow profiles show minor changes compared to the original unclamped configuration. Besides the velocity reduction at the outer wall of the left section flow is slightly increased in the right section.

Flow at the aneurysm site is shown in Figure 7. After ligation of LVA, the flow field is disturbed compared with the original unligated configuration. In contrast, ligation of the RVA causes flow similar to the flow in the original configuration. Flow reduction after ligation but with constant total flow was quantified near the aneurysm remnant region as marked in Figure 7. At the region of interest the in-plane velocity magnitude was significantly reduced (p<0.05, paired Wilcoxon test using SPSS 19, IBM Corp., Armonk, NY, USA) at the peak systole flow (t3) and during deceleration phase (t4) or approximately during 70% of the period. Figure 8 shows near remnant velocity values measured for the four time steps. Exemplary, at time step t3the mean ± standard deviation (s.d.) of flow was significantly reduced from 56.8 ± 11.6 mm/s (unligated) to 44.3 ± 18.7 mm/s (ligated LVA) and 38.6 ± 10.3 mm/s (ligated RVA). Flow reduction due to RVA clipping was also significant (p<0.05) relative to the flow generated by LVA clipping.

Near remnant velocity magnitude values at four characteristic time steps t1-t4 with the total in vivo flow rate of 133 ml/min in all three configurations: original unclamped – black circles ligated LVA – black squares and ligated RVA – black triangles. Values are shown without standard deviation (s.d.) bars for a better visibility. All s.d. ranged between 10 and 20 mm/s. Flow rate curve is added for a better data representation.


The case of a post-coiling aneurysm remnant at a fenestrated basilar artery was investigated. This seems to be a rare case. Wollschlaeger et al found more than 5% of fenestrated cases in a post-mortem study (20). In contrast, angiography studies show only a prevalence of about 1% of this anatomic variant. This is explained by the challenge to identify fenestrations in angiography data. Nevertheless, there are quite a few published cases (21-22-23-24-25). A case of this type was also a part of the Virtual Intracranial Stenting Challenge 2009 (26). It seems that fenestration promotes the genesis of aneurysms: Tsuei et al found aneurysms in 8 of 32 patients (25%) with the fenestrated vertebrobasilar junction (25). Another study, however, found aneurysms in only 7% of fenestrated basilar arteries (27).

Tsuei et al found no difference in hemodynamics of pre- and post-coiled geometries (25). Unfortunately, they do not describe the flow in detail. As we found, flow in a fenestrated basilar artery is a complex 3D disturbed flow, which is often associated with vessel wall remodeling, atherosclerosis, and aneurysm formation (28). In contrast, flow at a normal vertebrobasilar junction is described as parallel laminar flow in 80% of cases or as flow with secondary flow features (spiral flow), but without recirculation zones (29, 30). Consequently, it can be speculated that disturbed flow formed by a vortex as found in our case is associated with a degenerative, irregular shape of the RFVS vessel segment and with aneurysm development. Summarizing, hemodynamics and treatment of fenestrated artery aneurysms are of great interest.

Studies indicate that high wall shear stress associated with high velocities at the aneurysm neck is associated with aneurysm recanalization (31, 32). Hence, lowering of flow velocity near the neck or the remnant of an aneurysm due to post-ligation changes of hemodynamics (flow profiles or flow rate divisions at bifurcations) or due to reduction of total flow might reduce the recanalization rate. In our case, flow reduction near the aneurysm remnant was achieved by both ligation options (LVA or RVA ligation) even when total flow was not reduced. However, the impact of RVA ligation was significantly higher than that of LVA ligation. Flow reduction near the aneurysm remnant due to ligation of the LVA or RVA seems to result from a combined effect of the vortex preexisting at the junction and the curvature of VAs. Such an effect of flow manipulation cannot be predicted without extensive flow investigations.

A further possible effect of therapeutic interventions is the impact of altered hemodynamics on endothelial cells. Substantial flow reduction diminishes endothelial wall shear stress, thus decreasing local nitric oxide production (33). Nitric oxide inhibits platelet adherence and aggregation as well as leukocyte adhesion and infiltration (34). Therefore, velocity reduction at the aneurysmal neck might promote thrombus formation or inflammation.

The reliability of the proposed in vitro pipeline to improve the hemodynamic effect of ligation therapy was successfully shown in this case of a coiled fenestrated basilar artery aneurysm. The specific results of this hemodynamic study cannot be transferred to other cases of fenestrated basilar artery aneurysms, but the proposed in vitro study can be done on any other vessel configuration.

We only investigated the neurosurgical treatment option of Hunterian ligation. The definite treatment decision has not been made in this patient, who is currently in clinical and angiographical follow-up. The same pipeline can be used to study any other endovascular treatment option, for instance, like balloon occlusion or implantation of a flow diverting stent.

Alternatively, in silico pipeline using computational fluid dynamics tools could be applied (8, 26). Furthermore, a broad range of alternative experimental techniques (particle tracking velocimetry, laser Doppler velocimetry, and magnetic resonance imaging) was used in different studies of cerebral aneurysm hemodynamics (35-36-37-37). Comparing in vitro and in silico pipeline there are differences in technical challenges, technique availability, and expenditure of time and money. Nevertheless, we consider both techniques as basically equivalent.

Study limitations

This study has some limitations. The patient-specific flow waveform was unknown. Therefore, we used literature data. It should be noted that Hunterian ligation may also affect the flow waveform. It was shown that the waveform affects time-dependent flow parameters, whereas time-averaged flow parameters are independent (38). We found different flow conditions during different phases (compare results for time steps t1 and t3) of the heart cycle. This can be affected by the waveform used. The test fluid in our experiments exhibits Newtonian behavior, whereas blood is a non-Newtonian fluid (39). Use of non-patient-specific flow wave form and rate, and a Newtonian blood assumption may affect the accuracy of the calculated flow field. However, the geometry of the vessels is by far the most important factor affecting hemodynamics (28, 40).

Standard in-plane PIV was used in our study. Quantitative results obtained by PIV are limited by the spatial and temporal resolution of the respective techniques, including particle size and time intervals used for the data averaging (41). The PIV setting used is a compromise between data resolution and time costs of experiments, allowing an accurate velocity field assessment (17-18-19, 41).

Max-min images show regions (stripes) without tracers (see Fig. 5a). These are non-illuminated regions (cast shadows) due to small bubbles inside the wall of the silicone model (fabrication artefacts). We measured that the maximal width of a stripe without speckle, which is necessary for the PIV algorithm, was only 8 px. This does not affect results obtained with a 32 × 32 px interrogation area.

Alternative to the PIV technique used, other visualization techniques including LDV, holographic PIV, point-based 3D volumetric measurements or Tomo-PIV are possible (42-43-44). The measurement technique used is also a compromise between data resolution, technique availability, and time costs of experiments, including post-processing.


As illustrated by the current PIV study, in vitro methods of biomedical engineering can be used to reliably describe hemodynamic phenomena in complex configurations of the cerebral vessel. The presented case demonstrates the relevance of patient-specific flow analysis for treatment decisions and the reliability of the proposed experimental pipeline. For Hunterian ligation as a therapeutic option in a fenestrated basilar artery, the results show the sensitivity of fenestra flow to inflow conditions, which are substantially affected by the selection of the artery to be ligated. With retained total flow, ligation of either of the vertebral arteries results in significant flow reduction at the aneurysm site. Hence, Hunterian ligation might be successfully used.


Financial Support: This study was supported by the German Research Foundation (DFG project KE 900/9-1).
Conflict of Interest: None.
  • 1. Brown RD. Unruptured intracranial aneurysms. Semin Neurol 2010; 30: 537-544 Google Scholar
  • 2. Lasheras J. The biomechanics of arterial aneurysms. Annu Rev Fluid Mech 2007; 39: 293-319 Google Scholar
  • 3. Darsaut TE.,Bing F.,Salazkin I.,Gevry G.,Raymond J. Testing flow diverters in giant fusiform aneurysms: a new experimental model can show leaks responsible for failures. AJNR Am J Neuroradiol 2011; 32: 2175-2179 Google Scholar
  • 4. Heros RC.,Morcos JJ. Cerebrovascular surgery: past, present, and future. Neurosurgery 2000; 47: 1007-1033 Google Scholar
  • 5. Polevaya NV.,Kalani MY.,Steinberg GK.,Tse VC. The transition from Hunterian ligation to intracranial aneurysm clips: a historical perspective. Neurosurg Focus 2006; 20: 1-7 Google Scholar
  • 6. Clarke M. Systematic review of reviews of risk factors for intracranial aneurysms. Neuroradiology 2008; 50: 653-664 Google Scholar
  • 7. Suarez JI.,Tarr RW.,Selman WR. Aneurysmal subarachnoid hemorrhage. N Engl J Med 2006; 354: 387-396 Google Scholar
  • 8. Shojima M.,Morita A.,Kimura T.,Oshima M.,Kin T.,Saito N. Computational fluid dynamic simulation of a giant basilar tip aneurysm with eventual rupture following Hunterian ligation. World Neurosurg ; : - Google Scholar
  • 9. Goubergrits L.,Schaller J.,Kertzscher U. Statistical wall shear stress maps of ruptured and unruptured middle cerebral artery aneurysms. J R Soc Interface 2012; 9: 677-688 Google Scholar
  • 10. Goubergrits L.,Thamsen B.,Berthe A. In vitro study of near-wall flow in a cerebral aneurysm model with and without coils. AJNR Am J Neuroradiol 2010; 31: 1521-1528 Google Scholar
  • 11. Alastruey J.,Parker KH.,Peiró J.,Byrd SM.,Sherwin SJ. Modelling the circle of Willis to assess the effects of anatomical variations and occlusions on cerebral flows. J Biomech 2007; 40: 1794-1805 Google Scholar
  • 12. Goubergrits L.,Weber S.,Petz Ch. Wall-PIV as a near wall flow validation tool for CFD: application in a pathologic vessel enlargement (aneurysm). J Vis (Tokyo) 2009; 12: 241-250 Google Scholar
  • 13. Cebral JR.,Castro MA.,Putman CM.,Alperin N. Flow-area relationship in internal carotid and vertebral arteries. Physiol Meas 2008; 29: 585-594 Google Scholar
  • 14. Kato T.,Indo T.,Yoshida E.,Iwasaki Y.,Sone M.,Sobue G. Contrast-enhanced 2D cine phase MR angiography for measurement of basilar artery blood flow in posterior circulation ischemia. AJNR Am J Neuroradiol 2002; 23: 1346-1351 Google Scholar
  • 15. Rutgers DR.,Blankensteijn JD.,van der Grond J. Preoperative MRA flow quantification in CEA patients: flow differences between patients who develop cerebral ischemia and patients who do not develop cerebral ischemia during cross-clamping of the carotid artery. Stroke 2000; 31: 3021-3028 Google Scholar
  • 16. Rayz VL.,Boussel L.,Acevedo-Bolton G. Numerical simulations of flow in cerebral aneurysms: comparison of CFD results and in vivo MRI measurements. J Biomech Eng 2008; 130: 051011- Google Scholar
  • 17. Adrian RJ. Twenty years of particle image velocimetry. Exp Fluids 2005; 39: 159-169 Google Scholar
  • 18. Grant I. Particle image velocimetry: a review. J Mech Eng Sci 1997; 211: 55-76 Google Scholar
  • 19. Westerweel J. Theoretical analysis of the measurement precision in particle image velocimetry. Exp Fluids 2000; 29: S003-S012 Google Scholar
  • 20. Wollschlaeger G.,Wollschlaeger PB.,Lucas FV.,Lopez VF. Experience and results with postmortem cerebral angiography performed as a routine procedure of the autopsy. Am J Roentgenol Radium Ther Nucl Med 1967; 101: 68-87 Google Scholar
  • 21. Albanese E.,Russo A.,Ulm AJ. Fenestrated vertebrobasilar junction aneurysm: diagnostic and therapeutic considerations. J Neurosurg 2009; 110: 525-529 Google Scholar
  • 22. Fujimura M.,Sugawara T.,Higuchi H.,Oku T.,Seki H. A ruptured aneurysm at the distal end of the basilar artery fenestration associated with multiple fenestrations of the vertebrobasilar system: case report. Surg Neurol 1997; 47: 469-472 Google Scholar
  • 23. Peltier J.,Gayet J-B.,Toussaint P.,Deramond H.,Le Gars D. Anevrisme termino-vertebral sur fenestration de l’artere basilaire. [Terminovertebral aneurysm arising from basilar artery fenestration.] [in French] Neurochirurgie 2006; 52: 52-56 Google Scholar
  • 24. Saatci I.,Cekirge HS.,Karcaaltincaba M. Endovascular treatment of kissing aneurysms at the fenestrated basilar artery. Case report with literature review. Surg Neurol 2002; 58: 54-58 Google Scholar
  • 25. Tsuei Y-S.,Matsumoto Y.,Ohta M.,Nakayama T.,Ezura M.,Takahashi A. Vertebrobasilar junction fenestration with dumbbell-shaped aneurysms formation: computational fluid dynamics analysis. Surg Neurol 2009; 72: S11-S19 Google Scholar
  • 26. Neugebauer M.,Janiga G.,Beuing O.,Skalej M.,Preim B. Computer-aided modelling of blood flow for the treatment of cerebral aneurysms. Biomed Tech (Berl) 2010; 55: 37-41 Google Scholar
  • 27. Sanders WP.,Sorek PA.,Mehta BA. Fenestration of intracranial arteries with special attention to associated aneurysms and other anomalies. AJNR Am J Neuroradiol 1993; 14: 675-680 Google Scholar
  • 28. Steinman DA. Image-based computational fluid dynamics modeling in realistic arterial geometries. Ann Biomed Eng 2002; 30: 483-497 Google Scholar
  • 29. Krijger JK.,Heethaar RM.,Hillen B.,Hoogstraten HW.,Ravensbergen J. Computation of steady three-dimensional flow in a model of the basilar artery. J Biomech 1992; 25: 1451-1465 Google Scholar
  • 30. Smith AS.,Bellon JR. Parallel and spiral flow patterns of vertebral artery contributions to the basilar artery. AJNR Am J Neuroradiol 1995; 16: 1587-1591 Google Scholar
  • 31. Li CH.,Wang SZ.,Chen JL. Influence of hemodynamics on recanalization of totally occluded intracranial aneurysms: a patient-specific computational fluid dynamic simulation study. J Neurosurg 2012; 117: 276-283 Google Scholar
  • 32. Luo B.,Yang X.,Wang S. High shear stress and flow velocity in partially occluded aneurysms prone to recanalization. Stroke 2011; 42: 745-753 Google Scholar
  • 33. Chiu JJ.,Chien S. Effects of disturbed flow on vascular endothelium: pathophysiological basis and clinical perspectives. Physiol Rev 2011; 91: 327-387 Google Scholar
  • 34. Davignon J.,Ganz P. Role of endothelial dysfunction in atherosclerosis. Circulation 2004; 109: III27-III32 Google Scholar
  • 35. Funamoto K.,Suzuki Y.,Hayase T.,Kosugi T.,Isoda H. Numerical validation of MR-measurement-integrated simulation of blood flow in a cerebral aneurysm. Ann Biomed Eng 2009; 37: 1105-1116 Google Scholar
  • 36. Liou T-M.,Li Y-Ch.,Juan W-Ch. Numerical and experimental studies on pulsatile flow in aneurysms arising laterally from a curved parent vessel at various angles. J Biomech 2007; 40: 1268-1275 Google Scholar
  • 37. Morino T.,Tanoue T.,Tateshima S.,Vinuela F.,Tanishita K. Intra-aneurysmal blood flow based on patient-specific CT angiogram. Exp Fluids 2010; 49: 485-496 Google Scholar
  • 38. Karmonik Ch.,Yen Ch.,Diaz O.,Klucznik R.,Grossman RG.,Benndorf G. Temporal variations of wall shear stress parameters in intracranial aneurysms—importance of patient-specific inflow waveforms for CFD calculations. Acta Neurochir (Wien) 2010; 152: 1391-1398 Google Scholar
  • 39. Fisher C.,Rossmann JS. Effect of non-newtonian behavior on hemodynamics of cerebral aneurysms. J Biomech Eng 2009; 131: 091004- Google Scholar
  • 40. Hoi Y.,Woodward SH.,Kim M.,Taulbee DB.,Meng H. Validation of CFD simulations of cerebral aneurysms with implication of geometric variations. J Biomech Eng 2006; 128: 844-851 Google Scholar
  • 41. Nanna JC.,Navitsky MA.,Topper SR.,Deutsch S.,Manning KB. A fluid dynamics study in a 50 cc pulsatile ventricular assist device: influence of heart rate variability. J Biomech Eng 2011; 133: 101002- Google Scholar
  • 42. Buchmann NA.,Atkinson C.,Jeremy MC.,Soria J. Tomographic particle image velocimetry investigation of the flow in a modeled human carotid artery bifurcation. Exp Fluids 2011; 50: 1131-1151 Google Scholar
  • 43. Graff EC.,Gharib M. Performance prediction of point-based three-dimensional volumetric measurement systems. Meas Sci Technol 2008; 19: 075403- Google Scholar
  • 44. Hinsch KD. Three-dimensional particle velocimetry. Meas Sci Technol 1995; 6: 742- Google Scholar



  • Biofluid Mechanics Laboratory, Charité – Universitätsmedizin Berlin, Berlin - Germany
  • Department of Neurosurgery, Helios Hospital Berlin-Buch, Berlin - Germany
  • Visualization and Data Analysis, Zuse Institute Berlin, Berlin - Germany
  • These authors contributed equally to this manuscript

Article usage statistics

The blue line displays unique views in the time frame indicated.
The yellow line displays unique downloads.
Views and downloads are counted only once per session.

No supplementary material is available for this article.