Determinants of left ventricular circumferential-radial shear strain

$286

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

6028 Ventricular-vascular

Tu, 08:45-09:00 (P17) coupling in numerical models o f cardiac

electromechanics R. Kerckhoffs 1, Q. Gu 2, J. Omens 1,3, A. McCulloch 1. 1Department of

Bioengineering, UC, San Diego, USA, 2Department of Structural Engineering, UC, San Diego, USA, 3Department of Medicine, UC, San Diego, USA Finite element (FE) models of the heart have been used to investigate regional myofiber strains and stresses during normal and abnormal function. To ensure a reasonable realistic afterload, two-element [1] and three-element windkessel models [2] have been coupled to FE models. Preload was prescribed by an (unphysiological) linear increase in diastolic pressure. To prescribe a physiological preload, a left ventricular FE model was coupled to a model of pulmonary venous return [3]. However, the thus far used pre- and afterload models do not ensure conservation of blood mass; the FE models were not embedded in the circulation. The aim of this study was to develop a method to couple FE models of the heart to lumped models of circulatory blood flow dynamics. We used three-element windkessel models coupled to a dog model 1 as a test case. The method is based on estimating ventricular cavity pressures [2]. In our method, pressure is estimated using ventricular compliances (dVLv/dpLv, dVRv/dpRv [ml/kPa]) and ventricular interaction coefficients (dVRv/dpLv, dVLv/dpRv [ml/kPa]) from FE and circulatory model by perturbation. The estimated pressure is prescribed in the FE model and circulatory model. The volume error (FE cavity volume minus circulatory cavity volume) is minimized using Newton iterations. After perturbation, pressures converged within 2 estimations, with a relative cavity volume error of <0.1%. FE interaction coefficients were relatively large compared to compliance: dVRv/dpLv =-4.35, dVLv/dpRv =-3.61, dVLv/dpLv=9.88, dVRv/dpRv=18.6 at enddiastole and dVRv/dpLv=-0.748, dVLv/dpRv = -0.659, dVLv/dpLv = 2.75, dVRv/dpRv = 4.48 at begin ejection. Advantages of this method over previous methods are: a. ventricular interaction taken into account, resulting in fast convergence b. independent of cardiac phase c. modular setup: FE model, pressure estimation algorithm and circulatory model separated d. easy application of pathologies (e.g. regurgitation, ventricular defects) e. easy embedding of FE model in closed circulation. References

[1] Usyk et al. Comp. Visual. Sci. 2002; 4: 249-257. [2] Bovendeerd et al. J. Biomech. 1992; 25:1129-1140. [3] Watanabe et al. B.iophys J. 2004; 87: 2074-2085. 4051

Tu, 09:00-09:15 (P17) coupling in a rat model o f reduced arterial compliance provoked by hypervitaminosis D and nicotine D. Jegger 1,2, R. da Silva 1, X. Jeanrenaud 3, M. Nasratullah 3, H. Tevaearai4, L.K. von Segesser 2, P. Segers 6, V. Gaillard 5, J. Atkinson 5, I. Lartaud 5, N. Stergiopulos 1. 1Laboratory ef Haemedynamics and Cardiovascular

Ventricular-arterial

Technology, EPFL; 2Dept. of Cardiovascular Surgery, 3Dept. of Cardiology, CHUV, Lausanne, Switzerland, 4Dept. of Cardiovascular Surgery, Inselspital, Bern, Switzerland, 5pharmacology Laboratory, Pharmacy Faculty, Nancy, France, ~Hydraulics Laboratory, Ghent University, Gent, Belgium Objective: Rodent models of isolated systolic hypertension (ISH) are rare. One exception is the vitamin D and nicotine (VDN) model, in which arterial calcification raises arterial stiffness and vascular impedance. A complete analysis of the effect of VDN on ventricular-arterial interaction is lacking. Methods: Wistar rats were treated with VDN (VDN group, n =9) and a control group (CTRL, n= 10) was included without the VDN. At sacrifice, invasive indexes of cardiac function were obtained using a conductance catheter. Aortic pressure and flow were measured to derive impedance and ventricular-arterial interaction. Results: VDN caused significant increases in systolic (138±6mmHg vs. 116±13mmHg, p<0.01) and pulse pressures (42±10mmHg vs. 26±4 mmHg, p < 0.01) with respect to CTRL. Arterial compliance decreased (0.12±0.08 ml/mmHg vs. 0.21 ±0.04 ml/mmHg CTRL, p < 0.05) and pulse wave velocity increased significantly (8.8±2.5 m/s vs. 5.1±2.0 m/s CTRL, p < 0.05). Elastic modulus rose in the VDN group (p<0.05). Preload recruitable stroke work and end-systolic elastance were both elevated in the VDN group thus decreasing the ratio of arterial elastance over end-systolic elastance (0.94±0.30 vs. 1.57±0.60 CTRL, p <0.05) and augmenting efficiency. Wave reflection was augmented in the VDN group, expressed by the increase in the wave reflection coefficient, A (0.63±0.06 vs. 0.52±0.05 CTRL, p<0.05), as well as the amplitude of the reflected pressure wave (13.3±3.1 mmHg vs. 8.4±1.0 mmHg CTRL, p < 0.05). Conclusions: VDN lead to development of ISH and provoked alterations in cardiac function, arterial impedance, arterial function, and ventricular-arterial interaction, which in many aspects are similar to effects of an aged and

Oral Presentations stiffened arterial tree. The VDN model may be a useful model to study the patho-physiological effects of increased arterial stiffness. 6492

Tu, 09:15-09:30 (P17)

Determinants of left ventricular circumferential-radial shear strain

P. Bovendeerd 1, S. Ubbink 1, T. Delhaas 2, T. Arts 3. 1Dept ef Biomedical Engineering, Eindhoven Univ. of Technology, Eindhoven, The Netherlands, 2 Dept of Physiology and 3 Dept of Biophysics, Maastricht University, Maastricht, The Netherlands Magnetic resonance tagging (MRT) is an attractive technique to assess local myocardial deformation. Understanding how myocardial tissue properties, cardiac fiber structure, and cardiac geometry determine cardiac deformation is important for the interpretation of the tagging data. Usually, MRT data are analysed in terms of circumferential strain [1], and inhomogeneities in circumferential strain are related to ischemia or conduction disorders. Circumferential-radial shear strain may contain additional information. However, while this strain component has been proposed to be related to the sheet structure of the myocardial wall [2,3], multi axial active stress generation [3] and myofiber orientation [4], the sensitivity to each of these factors is unknown. In this study we used a finite element model of left ventricular mechanics to investigate this sensitivity. In addition, we measured left ventricular wall shear strain in healthy volunteers with MRT, and investigated the extent to which shear strain in model and experiment could be matched. References

[1] [2] [3] [4]

Wyman et al. American Journal of Physiology 1999; 276: H881-H891. LeGrice et al. Circulation Research 1995; 77: 182-193. Usyk et al. Journal of Elasticity 2000; 61: 143-164. Bovendeerd et al. Journal of Biomechanics 1994; 27: 941-951.

7868 Tu, 09:30-09:45 (P17) Understanding and measuring flow in aortic stenosis with magnetic resonance imaging K. O'Brien 1, B. Cowan 2, R. Stewart 2, A. Young 1. 1Bioengineering Institute,

2 The University of Auckland, New Zealand Severity of aortic stenosis (AS) is the primary indicator for a patient to undergo aortic valve replacement (AVR). Tools are therefore required that provide an accurate assessment of AS severity and ventricular function to determine progression of disease and predict when AVR is required. Doppler and Two-dimensional echocardiography is currently used clinically to assess AS severity through characterising the aortic valve pressure gradient and the aortic valve area. The major limitation of Doppler is the restriction, of the limited anatomical windows, in aligning the beam parallel to the AS velocity jet. This can result in patients with severe AS to be misdiagnosed as having moderate or mild AS. Estimation of valve area is also critically dependent on the estimation of the left ventricular outflow tract diameter and is effected by poor quality trans-thoracic images. We acquired high resolution 3D velocity data before and after the aortic valve using phase contrast magnetic resonance imaging (MR). Ventricular performance was determined by fitting a mathematical model to the left ventricle throughout the cardiac cycle. These data can be used to test computational fluid dynamics (CFD) models, which can then be used to estimate the geometric parameters (position, shape, area) of a valve lesion to match the measured MR velocity data. The CFD models and velocity estimation methods can be validated using custom make MR flow phantoms that simulate the valve using an adjustable orifice plate. Results will be presented from the flow phantoms and preliminary analyses data from 60 patients. 7849

Tu, 11:00-11:30 (P19)

Arterial wall mechanics and remodeling using a constituent-based

approach N. Stergiopulos. Ecele Polytechnique F~d6rale de Lausanne (EPFL),

Hemodynamics and Cardiovascular Technology Laboratory (LHTC), Lausanne, Switzerland Earlier studies in experimental hypertension have shown that acute hypertension leads to wall remodeling, which, in general, aims in restoring mean wall (hoop) stress to control levels. One may thus postulate that wall remodeling should be such that local stresses are normalized. This postulate has not been yet thoroughly studied, mainly because precise knowledge of stresses acting on each wall constituent and in all parts of the wall is still a very difficult task. This requires a constituent-based modeling and analysis of the arterial wall. The three-dimensional biomechanical behavior of the vascular wall is best described by means of strain energy functions, which allow for the analysis of stresses over a wide range of deformations. We have earlier developed appropriate strain energy functions for the arterial wall [1]. The Zulliger et al. model uses a strain energy function, which accounts for the