Adjoint-based inverse analysis of windkessel parameters for patient-specific vascular models

A human circulatory system is composed of more than 50,000 miles of blood vessels. Such a huge network of vessels is responsible for the elevated pressure values within large arteries. As such, modeling of large blood arteries requires a full modeling of circulatory system. This in turn is computationally not affordable. Thus, a multiscale modeling of the arterial network is a necessity. The multiscale approach is achieved through, first, modeling the arterial regions of interest with 3D models and the rest of the circulatory network with reduced-dimensional (reduced-D) models, then coupling the multiscale domains together. Though reduced-D models can well reproduce physiology, calibrating them to fit 3D patient-specific Fluid Structure Interaction (FSI) geometries has received little attention. For this reason, this work develops calibration methods for reduced-D models using adjoint based methods. We also propose a reduced modeling complexity (RMC) approach that reduces the calibration cost of expensive FSI models using pure fluid modeling. Finally, all of the developed calibration techniques are tested on patient-specific arterial geometries, showing the power and stability of the proposed calibration methods.

[1]  Christian Reeps,et al.  A continuum description of the damage process in the arterial wall of abdominal aortic aneurysms , 2012, International journal for numerical methods in biomedical engineering.

[2]  Lieve Lanoye,et al.  Effect of an Abdominal Aortic Aneurysm on Wave Reflection in the Aorta , 2008, IEEE Transactions on Biomedical Engineering.

[3]  Frans N van de Vosse,et al.  Effects of wall calcifications in patient-specific wall stress analyses of abdominal aortic aneurysms. , 2007, Journal of biomechanical engineering.

[4]  A. Quarteroni,et al.  One-dimensional models for blood flow in arteries , 2003 .

[5]  Pablo J. Blanco,et al.  Assessing the influence of heart rate in local hemodynamics through coupled 3D‐1D‐0D models , 2010 .

[6]  S. Sherwin,et al.  Lumped parameter outflow models for 1-D blood flow simulations: Effect on pulse waves and parameter estimation , 2008 .

[7]  A. Quarteroni,et al.  On the coupling of 3D and 1D Navier-Stokes equations for flow problems in compliant vessels , 2001 .

[8]  Peng Zhao,et al.  Blood Flow in the Circle of Willis: Modeling and Calibration , 2008, Multiscale Model. Simul..

[9]  J Swedenborg,et al.  Biomechanical rupture risk assessment of abdominal aortic aneurysms: model complexity versus predictability of finite element simulations. , 2010, European journal of vascular and endovascular surgery : the official journal of the European Society for Vascular Surgery.

[10]  Wolfgang A. Wall,et al.  A computational strategy for prestressing patient‐specific biomechanical problems under finite deformation , 2010 .

[11]  Volker Gravemeier,et al.  A novel formulation for Neumann inflow boundary conditions in biomechanics , 2012, International journal for numerical methods in biomedical engineering.

[12]  Charles A. Taylor,et al.  Tuning Multidomain Hemodynamic Simulations to Match Physiological Measurements , 2010, Annals of Biomedical Engineering.

[13]  Berend E. Westerhof,et al.  The arterial Windkessel , 2009, Medical & Biological Engineering & Computing.

[14]  W A Wall,et al.  Impact of calcifications on patient-specific wall stress analysis of abdominal aortic aneurysms , 2010, Biomechanics and modeling in mechanobiology.

[15]  S. Sherwin,et al.  Modelling the circle of Willis to assess the effects of anatomical variations and occlusions on cerebral flows. , 2007, Journal of biomechanics.

[16]  Wolfgang A Wall,et al.  Structured Tree Impedance Outflow Boundary Conditions for 3D Lung Simulations , 2010, Journal of biomechanical engineering.

[17]  Ali Nadim,et al.  On deriving lumped models for blood flow and pressure in the systemic arteries. , 2004 .

[18]  P. Nithiarasu,et al.  A 1D arterial blood flow model incorporating ventricular pressure, aortic valve and regional coronary flow using the locally conservative Galerkin (LCG) method , 2008 .

[19]  W A Wall,et al.  Prestressing in finite deformation abdominal aortic aneurysm simulation. , 2009, Journal of biomechanics.

[20]  A Comerford,et al.  A stable approach for coupling multidimensional cardiovascular and pulmonary networks based on a novel pressure‐flow rate or pressure‐only Neumann boundary condition formulation , 2014, International journal for numerical methods in biomedical engineering.

[21]  Alfio Quarteroni,et al.  Multiscale modelling of the circulatory system: a preliminary analysis , 1999 .

[22]  Wolfgang A. Wall,et al.  Coupling strategies for biomedical fluid–structure interaction problems , 2010 .

[23]  Charles A. Taylor,et al.  Outflow boundary conditions for three-dimensional finite element modeling of blood flow and pressure in arteries , 2006 .

[24]  M Ismail,et al.  Coupled and reduced dimensional modeling of respiratory mechanics during spontaneous breathing , 2013, International journal for numerical methods in biomedical engineering.

[25]  M. Olufsen,et al.  Dynamics of cerebral blood flow regulation explained using a lumped parameter model. , 2002, American journal of physiology. Regulatory, integrative and comparative physiology.

[26]  S. Sherwin,et al.  One-dimensional modelling of a vascular network in space-time variables , 2003 .

[27]  Christian Reeps,et al.  The impact of model assumptions on results of computational mechanics in abdominal aortic aneurysm. , 2010, Journal of vascular surgery.

[28]  Gerhard A. Holzapfel,et al.  Modeling Plaque Fissuring and Dissection during Balloon Angioplasty Intervention , 2007, Annals of Biomedical Engineering.

[29]  A. Avolio,et al.  Multi-branched model of the human arterial system , 1980, Medical and Biological Engineering and Computing.

[30]  W. Wall,et al.  Truly monolithic algebraic multigrid for fluid–structure interaction , 2011 .

[31]  W. Wall,et al.  A Comparison of Diameter, Wall Stress, and Rupture Potential Index for Abdominal Aortic Aneurysm Rupture Risk Prediction , 2010, Annals of Biomedical Engineering.

[32]  Spencer J. Sherwin,et al.  Computational modelling of 1D blood flow with variable mechanical properties and its application to the simulation of wave propagation in the human arterial system , 2003 .

[33]  M. Olufsen,et al.  Numerical Simulation and Experimental Validation of Blood Flow in Arteries with Structured-Tree Outflow Conditions , 2000, Annals of Biomedical Engineering.

[34]  M L Raghavan,et al.  Toward a biomechanical tool to evaluate rupture potential of abdominal aortic aneurysm: identification of a finite strain constitutive model and evaluation of its applicability. , 2000, Journal of biomechanics.

[35]  Charles A. Taylor,et al.  Intracranial and abdominal aortic aneurysms: similarities, differences, and need for a new class of computational models. , 2008, Annual review of biomedical engineering.