Fluid-structure interaction modelling for assessing dilatation of ascending aorta

Motivation

Within CARDIOPROOF Siemens is interested in validating coupled computational models of blood flow and wall deformation for assessing the role of haemodynamics, wall mechanics and anatomy in dilatation of ascending aorta. Aortic arch dilatation occurs frequently and at a young age in patients with aortic disease, such as aortic valve disease (AVD) and aortic coarctation (CoA) specifically addressed in the project. Moreover, independently of the type of aortic disease, aortic dilation is a significant predictor of dissection and rupture (acute cardiovascular events leading to death if untreated), thus evolution of aortic arch dilation is important for prognostics.

nsl-4
Information on pressure and velocity of blood flow in the human cardiovascular system can be decisive for clinical evaluations (initial and post-procedural) and procedure planning. For example, the severity of the cardiovascular diseases targeted in this project can be assessed by intraluminal pressure gradients (Currie, et al. 1985). Furthermore, local arterial wall properties are important indicators for the development and assessment of cardiovascular diseases. More specifically, increased aortic arch stress (of both tensile and shear type) plays a role in pathogenesis of aortic wall remodelling and dilation. However, to date local aortic pressure measurements are only acquired invasively through catheterization, and local arterial wall properties are not directly measurable. Within CARDIOPROOF, we investigate the feasibility of combining non-invasive magnetic resonance imaging (MRI) with computational models of blood flow and wall deformation to derive both pressure and local arterial wall properties. The outcome of our research could provide cardiologists with enriched data from state-of-theart non-invasive imaging to support them in their complex decisions.

Computing the pressure difference field from phase contrast MRI

 In the first part of the project we focused our efforts on a method to compute relative pressures within the aorta non-invasively from MRI. To this end, we developed a twostep approach. In the first step, the blood velocity field acquired by phase contrast MRI needs to be reconstructed as it is subject to measurement errors. In particular, it is important to regularize the measured data to ensure that the corrected blood velocity field complies with the law of physics such as incompressibility and boundary conditions at the vessel walls. In the second step, the reconstructed velocities can be used to derive the timevarying 3D relative pressure within the aorta using Poisson’s equation. Together with our clinical partners, we are now in the process of validating the approach by comparing the simulation results with invasive measurements. This will extend our understanding of the applicability and the limits of the approach in a real clinical setting. Exemplar simulation results are shown in the figure  1.

Outlook: Estimating regional mechanical properties of the vessel wall

More recently we started investigating an automated method to compute patientspecific local and regional arterial wall properties solely using information from noninvasively acquired MRI data of the aorta. The framework is based on a fluid-structure interaction (FSI) blood flow model, whose parameters are estimated through an advanced personalization procedure to ensure high goodness of fit between the computational output and the patient-specific measurements. The computational results are being integrated in the same clinical prototype as shown above to facilitate interpretation and validation by the clinical users.

by Olivier Ecabe- SIEMENS

For more information, see  “Newsletters- ISSUE 2