$274 Journal o f Biomechanics 2006, Vol. 39 (Suppl 1)

model. This suggests that AAA rupture risk classification based on centreline curvature and torsion could be used in conjunction with parameters such as peak aneurysm diameter, sac volume and asymmetry to improve patient management.

References [1] Y. Papaharilaou, J. Ekaterinaris, E. Manousaki and A. Katsamouris. A decoupled

fluid structure approach of estimating wall stress in abdominal aortic aneurysms. Journal of Biomechanics, 2006; in print.

[2] L. Antiga and D. Steinman. Robust and objective decomposition and mapping of bifurcating vessels. IEEE Trans Med Imaging. 2004 Jun; 23(6): 704-13.

5552 Mo, 11:15-11:30 (P8) Experimental and numerical studies on physio logical f low behav iour in an asymmetr ic model o f abdominal aor t ic aneurysm E. Galliard, V. Deplano, O. Boiron, E. Bertrand. IRPHE UMR 6594, Equipe de Biom6canique Cardiovasculaire, Marseille, France

An abdominal aortic aneurysm (AAA) is a localized dilatation of the abdominal aorta downstream from the renal arteries including sometimes the lilac bifur- cation. Rupture is the main complication of AAAs, in 80% of cases, it leads to death. The knowledge of the local haemodynamic, including velocity, vorticity, shear and pressure distributions, may be used to improve medical diagnosis on AAA rupture. Numerical and experimental studies are therefore carried out in a three dimen- sional asymmetric model of AAA with or without lilac bifurcation, to analyse the behaviour of physiological flows in AAA. Velocity measurements are performed using particle image velocimetry (PIV). In addition, a finite volume method [1] is used to perform three-dimensional unsteady numerical simulations. Different Womersley parameter values and Reynolds number values are used to assess the parameters affecting the flow behaviour. These parameters model a normal physiological flow rate, a moderate exercise flow rate and an intensive exercise flow rate. For the first time, rigid walls versus compliant ones and lilac bifurcation model downstream from AAA versus straight artery model have been experimentally investigated to analyse both the compliance and the bifurcation influences on AAA flow behaviour. Compliant wall mechanical behaviour is characterized using a classical traction bench [2]. The secondary flow patterns and more particularly the vortices trajectories and their impact on the distal AAA wall are found to be highly dependent on the flow waveforms, the wall behaviour and on the lilac bifurcation presence. These results can help to improve medical diagnosis on AAA rupture. In a future work, we would like to numerically model the thrombus formation to better represent the pathological condition.

References [1] Fluent Software, Fluent Inc., 1998. [2] J. Polansky, O. Boiron, V. Novacek. Identification of viscoelastic properties of

artificial materials simulating vascular wall, Computer Methods in Biomechanics and Biomedical Engineering 2005; Vol. 8, $1.

5187 Mo, 11:30-11:45 (P8) Rupture mechanisms in c i rculatory system vascular tissue H.W. Haslach Jr.. Department of Mechanical Engineering, University of Maryland, USA

Rupture of circulatory vascular tissue under non-impact loading has been attributed to disparate mechanisms. The cause of aneurysm rupture has been postulated to be stress exceeding a uniaxial limit in saccular or abdominal aortic aneurysms, a limit-point instability such as is observed in expanding rubber balloons, dynamic resonance, and large radius or curvature. Excessive strain has been postulated as the cause of rupture of small veins during shock wave lithotripsy. The presence of thrombus or plaque is thought to respectively lower wall stress or induce stress concentrations. Nonlinear dynamics, not resonance, can explain the bruit, a detectable sound due to the aneurysm wall vibration at a frequency above that of blood pressure (Haslach 2002), so dynamics may influence rupture. Uniaxial tensile tests of flat strips of bovine artery tissue have shown that rupture begins in the intimal layer, as expected. The fracture surface shows protruding fibers of different length due to fiber pull-out from the matrix or fiber fracture. Uniaxial fatigue tests examine whether dynamic loading may raise the local temperature above body temperature and threaten the reducible collagen fibril cross-links. The hypothesis presented is that rupture causing functional failure primarily depends on the behavior of the mesoscale structures of collagen fibers. Dynamic loading influences this behavior once damage has occurred. Previous degeneration may be partially due to the cyclic stress and strain effect on the cells; stress may do more than pull apart bonds as implicitly assumed in earlier models. Once a crack initiates, the crack may grow due to fatigue. A model for vascular tissue rupture is biaxial non-linear viscoelastic and accounts for the dynamic behavior of the composite structure, in particular the collagen fibers.

Oral Presentations

References H. W. Haslach Jr. (2002). A nonlinear dynamical mechanism for bruit generation

by an intracranial saccular aneurysm. Journal of Mathematical Biology 45(5): 441-460.

H. W. Haslach Jr. (2005). Nonlinear viscoelastic, thermodynamically consistent, models for biological soft tissue. Biomechanics and Modeling in Mechanobiology 3(3): 172-189.

5454 Mo, 11:45-12:00 (P8) Relat ionship between growth rate and maximum wall s t ress in abdominal aort ic aneurysms

J.H. Leung 1 , A. Wright 2, N. Cheshire 2, S.A. Thorn 3, A.D. Hughes 3, J. Crane 2, X.'~ Xu 1 . 1 Department of Chemical Engineering, South Kensington Campus, UK, 2 Vascular Surgery and Radiology, 2NHLI, International Centre for Circulatory Health, St Mary's Hospital & Imperial College London, UK

This study explores the relationship between wall mechanical stress and the expansion rate of abdominal aortic aneurysms (AAA). Weakening of the wall by enzymatic protease enhances aortic dilatation. It is known that cells along the aortic wall can respond biochemically to mechanical stress [1]. Four AAA patients with initial aneurysm diameters ranging from 3.3 mm to 6 mm had 1 to 4 follow-up scans over a period of up to 45 months. Patient-specific models were reconstructed based on CT images acquired from each scan and wall stress was calculated using a finite element package ADINA. Predicted maximum wall stresses were set in relation to their corresponding growth rates (mm/year). As suggested in previous studies [2,3], there is no simple correlation between the peak wall stress in AAA and its maximum diameter. Our new findings are that AAAs with low peak wall stress (under 350kPa) have lower growth rates (0 to 0.33 mm/year) and AAAs with high peak stress level (above 350kPa) have relatively high growth rates (0.73-1.1 mm/year). The results also show that for a given patient with a growing aneurysm the maximum wall stress increases with the increase in its expansion rate and reduces when the growth rate slows down. This suggests that mechanical stress may affect the pathology of the degrading aneurysm wall.

References [1] Nakahashi T.K., et al. Flow loading induces macrophage antioxidative gene

expression in experimental aneurysms. Arterioscler Thromb Vasc Biol. 2002; 22(12): 2017-22.

[2] Fillinger M.E, et al. Prediction of rupture risk in abdominal aortic aneurysm during observation: wall stress versus diameter. J Vasc Surg. 2003; 37(4): 724-32.

[3] Raghavan M.L., et al. Wall stress distribution on three-dimensionally recon- structed models of human abdominal aortic aneurysm. J Vasc Surg. 2000; 31 (4): 760-9.

5791 Mo, 12:00-12:15 (P8) Reduct ion o f post-operat ive abdominal aor t ic graft pressures

T.M. McGIoughlin 1, L.G. Morris 2, T.P. O'Brien 1. 1Centre for Applied Biomedical Engineering Research and Materials and Surface Science Institute, Department of Mechanical and Aeronautical Engineering, University of Limerick, Limerick, Ireland, 2Department of Mechanical Engineering, Galway Mayo Institute of Technology, Galway, Ireland

Introduct ion: Abdominal aortic aneurysm (AAA) is an irreversible dilation of the abdominal aorta. It affects up to 5% of males over the age of 55. If the aneurysm goes unnoticed or is untreated it may rupture. Traditional treatment involves removal of the diseased segment of the aorta and replacement with a graft fabricated from synthetic material such as polyester or ePTFE.Such an operation involves problems such as large blood loss, prolonged cross clamping of the aorta (increasing blood pressure), large risk of complications and long recovery periods. Recent research has suggested that the introduction of a graft may give rise to an increase in blood pressure and cause an associated rise in cardiac load. The objective of this study was to examine the influence of new geometrical features on aortic pressure measurements in a range of in vitro models. Materials and Methods: Wave reflections, aorto/iliac area ratio and graft stiffness are all believed to play a role in increasing aortic blood pressure. In order to assess the influence of these parameters on aortic pressure an experimental investigation was conducted. A computer controlled piston pump was used to generate a physiologically realistic aortic flow rate in a range of compliant silicone rubber models. These included models based on currently used AAA graft geometries and on a new tapered graft which is under development. Pressure measurements were obtained using a 0.5 mm pressure catheter at a location 5cm proximal to the lilac bifurcation in the model. Results: The maximum and minimum pressures were recorded for ten sam- ples for each model. The novel geometry was found to reduce both the peak pressure and the minimum pressure by approximately 9mm of Hg when compared with conventional grafts.

Track 14. Cardiovascular Mechanics

This reduction in aortic blood pressure may have a significant clinical benefit in the treatment of AAA and this will be discussed.

5254 Mo, 12:15-12:30 (P8) Pulsatile movement of the zenith aortic stent-graft

B.A. Howell 1 , D. Saloner 2, J.M. LaBerge 2, T.A.M. Chuter 2. 1Department ef Vascular Surgery, UCSF, San Francisco, USA, 2Department of Radiology, UCSF, San Francisco, USA

Objective: To measure the pulsatile movement of a bifurcated stent-graft, as the basis for pre-clinical durability testing. Method: We performed high resolution cine-fluoroscopy in 39 patients imme- diately following abdominal aortic aneurysm repair with a Zenith stent-graft and at 1, 12, 24, and 36 months of follow-up. We compared systolic and diastole images to assess the movement of landmarks on the stent-graft in the pararenal aorta, in the neck, in the aneurysm, and at both ends of each limb. Two types of movement were observed: movement of the entire stent body, or pulsatile translation (PT), and expansion/contraction of the stent body, or pulsatile diameter change (PDC). Results: Immediately after implantation, the mid-portion of the trunk showed the greatest pulsatile diameter change (PDC) from systole to diastole, but this segment also had the greatest decline in PDC over the first month of follow-up (p=0.03). One month after stent-graft insertion, the mean PDC of the mid-aneurysm stent was less than 0.6% of its diastolic diameter. The PDC of the paranrenal stent and the stent within the neck also declined rapidly during the first month, but from a lower initial levels. After one month, PDC remained below 1.5% of the diastolic diameter at all stent landmarks. Pulsatile translation (PT) did not change with duration of follow-up. If anything, there was a non-significant trend towards an increase. The distal ends of the limbs had consistently lower PT (p <0.005) than the body of the graft. The difference between the relative positions of the ends of the graft limbs and the main body of the graft in systole and diastole produced pulsatile limb bending. Conclus ions: All portions of the stent-graft ceased expanding and contracting to any significant degree within the one month of implantation. The observed PDC in these areas (approximately 1.0%) is far less than the 5.0% cyclical strain typically used as the basis pre-clinical finite element analysis and accelerated durability testing.

14.1.3. Endevascular Aneurysm Repair

7654 Mo, 14:00-14:30 (P11) Wall s t ress analys is can predict the success of endovascular aneurysm repair

A. Chaudhuri 1 , P. Buxton 1 , L.E. Ansdell 2, M. Adiseshiah 1 , A.J. Grass 2. 1 Vascular Endovascular Unit, University College London Hospitals, London, UK, 2Department of Civil Engineering, Chadwick Building, University College London, London, UK

Objective(s): To assess peak wall stress changes after successful endovas- cular aneurysm repair (EVAR) and also failure, represented here as type 1/11 endoleakage. Methods: A 25 mm Talent endovascular stent-graft was deployed in a life-like non-axisymmetric latex abdominal aortic aneurysm model, which was incor- porated into a pulsatile flow unit. This was surrounded by thrombus analogue. Strain gauges were placed at the neck (n =2 3), inflection point (n =4 3) and maximum anteroposterior diameter (n =4 3) resulting in 24 functional output channels. The arterial pressure settings used were 130/90 and 140/100 mmHg, termed the low and high setting respectively. Strain readings were obtained at 10Hz over 30 seconds using a data logger before and after endograft deployment and after simulation of type I and II endoleaks. Stress was derived from its relationship with Young's modulus (E=5.151872Nmm-2). Peak wall stresses were statistically analysed using ANOVA in Minitab 13. Results (low/high settings): Peak stress was highest anteriorly and poste- riorly at the inflection point (394.69 (SD 218.1) 10-4/715.39 10-4N/cm -2 (SD 230.32), p<0.001, and 373.61(SD 207.24) 10-4/1053.32 (SD 347.01)

10-4Ncm -2, p<0.001). Type I endoleakage increased sac wall stresses (27.48-268.96/25.01-289.30 10-4Ncm -2, p<0.001), though these were equivocal at the neck and inflection point. Type II endoleakage also in- creased peak stresses (3.93-305.32/3.39-112.58 10 -4 Ncm -2, p < 0.001) though some reductions were noted at the inflection point. However, stresses produced by type I endoleakage were higher than that caused by a type II endoleak (p <0.001). Conclus ions: The therapeutic effect of EVAR is mediated by reduction of peak wall stresses. Type I endoleakage causes an increases in peak wall stress that may add to rupture risk. Stress reductions following type II endoleakage question whether this type needs treatment at all. This needs further validation in vive, which if successful, may allow non-invasive biomechanical post-EVAR AAA monitoring using CT-derived data.

14.1. Aneurysms - Endovascular Aneurysm Repair $275

7486 Mo, 14:30-14:45 ( P l l ) An in vitro assessment of Ancure and Zenith abdominal aortic aneurysm stent-graft devices under physio logical condi t ions

A. Callanan 1 , D. Kelly 1 , L. Morris 2, M. Walsh 1 , T. McGIoughlin 1 . 1Dept. Mechanical and Aeronautical Engineering, Centre for Biomedical Engineering Research (CABER) University of Limerick and MSSI, Ireland, 2Dept. Mechanical and Industrial Engineering, Galway Mayo Institute of technology, Galway, Ireland

Abdominal aortic aneurysms (AAA) are a major cause of death in the western world. Endovascular aneurysm repair (EVAR) is a less invasive procedure then the surgical method. Unfortunately stent graft devices have had many reported failures such as migration and endoleaks. Migration is one of the common causes and clinically most serious of the failure modes in AAA stent graft devices. The primary reason for this investigation is the assessment of prosthetic fixation, which is a vital to the long-term success of AAA Stent graft devices. Idealized and realistic AAA models were fabricated from silicone using an injection molding technique. A computer controlled piston pump was used to simulate a range of physiological loading conditions. Two commercial stent graft devices, the Ancure (Guidant) and Zenith (Cook) were tested in this in vitro setup. Static dislodgement loads was applied to both stent graft devices in conjunction to pulsatile flow and non-pulsatile flow conditions. A dislodgement force of 14N and 4.8N was found for both the Ancure and Zenith stent grafts respectively under no flow conditions A reduction of around 60% and 70% for the dislodgement force was found for the resting and exercise flow conditions respectively. Heart rate, pulse pressure and aortic properties all influence the device performance, and in particular the long-term attachment of AAA stent grafts. The Zenith stent graft system has a greater dislodgement load due to its higher radial force and supra renal fixation. The increase in pressure due to variations in the physiological type pulses causing migration at a decreased dislodgement load. Knowledge of physiological type flows and pulse pressure may prove vital in the improvement of performance of patient specific devices by aiding the selection of stent graft type.

7457 Mo, 14:45-15:00 ( P l l ) A computational model for endovascular graft sizing in abdominal aortic aneurysms

'~ Papaharilaou 1,3, J.A. Ekaterinaris 1,2, A.N. Katsamouris 3. 1Institute ef Applied and Computational Mathematics, Foundation for Research and Technology-Hellas, Heraklion, Greece, 2Dept. of Mechanical and Aerospace Engineering, University of Patras, Greece, 3Division of Vascular Surgery, Medical School, University of Crete, Greece

The sizing of endovascular grafts (EVG) for abdominal aortic aneurysm repair is a non-trivial procedure that often includes errors that compromise the quality of fit of the device and cause post-operative complications such as endoleaks and migration. In endovascular aneurysm repair (EVAR) secure device fixation at the proximal neck is critical for the long term success of the procedure. However, this often requires fitting a cylindrically shaped device to a tortuous conically shaped vessel lumen. This shape mismatch is typically addressed by oversizing the EVG a solution, however, that leads to a highly non uniform post operative stress distribution on the proximal neck wall and an equally non uniform device fixation strength. This in turn can lead to poor device post operative performance and arterial wall damage due to localized excessive tissue strain. Aim of this work is to model the fit of the EVG pre-operatively to predict the post-operative wall stress distribution in order to optimize device sizing. The EVG is modeled as a perfectly elastic isotropic material, the aortic wall as a hyperelastic isotropic material with zero residual stress, and the intraluminal thrombus as an elastic isotropic material [1]. Based on the known fully extended geometry of the EVG and the CT imaged 3D reconstructed aneurysmal true lumen geometry the initial stress distribution of the EVG that just fits in the AAA lumen is computed. This result is then prescribed as the initial EVG stress condition in the structural stress analysis of the EVG AAA assembly. The distribution of stress and frictional fixation force were computed. An initial validation of the proposed EVG fit modeling approach was achieved by comparing the model predicted fit geometry with the in vivo CT imaged 2-month post-operative EVAR geometry. Good agreement between the model fit and the post-operative fit was found.

References [1] Y. Papaharilaou, J. Ekaterinaris, E. Manousaki and A. Katsamouris. A decoupled

fluid structure approach of estimating wall stress in abdominal aortic aneurysms. Journal of Biomechanics, 2006; in print.