Dynamics of blood flow in the heart and large arteries

Investigation of blood flow using high-performance computing

The CompHem group focuses on the development and use of high-performance numerical tools to extract key information on blood flow in different geometrical configurations and for various applications. The main topics are thus:

(i) investigation of the flow in carotid bifurcations to understand the role of haemodynamic factors on atherosclerosis formation;

(ii) study of the hydrodynamic instabilities initiating laminar-to-turbulent transition downstream of bioprosthetic aortic valves and in different geometries of aortas; 

(iii) comprehensive modelling of the biomechanical and haemodynamic function of diseased mitral valves to help in the design of an optimised surgical repair procedure. 

The wide range of available tools, which are continuously developed and optimised to work on high-performance-computing (HPC) architectures allows us to perform simulations of fluid-structure interaction (FSI) problems, data-driven flow simulations and GPU-distributed large-scale simulations. The extracted in silico information is compared to in vivo 4D Flow Magnetic Resonance Imaging (MRI) measurements and to in vitro results obtained from three-dimensional particle image velocimetry in order to validate numerical models and to give clinical relevance to the obtained numerical results. 

Current contributors: Ali Mokhtari (PhD student) (i), Karoline-Marie Bornemann (PhD student) (ii), Valérie Kulka (PhD student) (iii).

Former members: Dr Pascal Corso - Dr Barna Becsek - Dr Dario De Marinis - Dr Hadi Zolfaghari.

Projects: PASC AV-FLOW (2014-2017), PASC HPC-PREDICT (2017-2021), SNSF-DFG Carotis (2021-2024), PASC MitrAccel (2021-2024).

[Text written by Dr P. Corso]

Optical methods and flow loops for the study of blood flow in the aorta

The experimental haemodynamics group makes use of in vitro left heart mock loops and silicone phantoms of the aorta to investigate the impact of various parameters on the flow features in the aorta and valve kinematics for different prosthetic or ex vivo heart valves. We aim at: (i) investigating the three-dimensional flow downstream of prosthetic heart valves inside aortic phantoms made of silicone to characterise stagnation regions, shear stresses and turbulence(A); (ii) investigating the influence of valve stiffness on valve opening and haemodynamics(B); and (iii) investigating the impact of aortic morphology and compliance on valve kinematics and haemodynamics(C).

Current projects and methods are summarised as follows: 
A tomographic particle image velocimetry (Tomo-PIV) setup is coupled with a pulse replicator to extract three-dimensional information on the flow behind heart valve prostheses. In this project(A), we study the performance of prosthetic valves by analysing the three-dimensional flow inside silicone phantoms of the aorta to highlight regions with low-velocity blood flow and quantify shear stresses. Moreover, the turbulent wakes generated by the valves are characterised using a spectral analysis. We investigate and compare different types of valves including surgically implanted (biological and mechanical) valves and valves implanted through a catheter. 

We designed a model of aortic valve stenosis using porcine valves that we treat to make them stiffer and then externally measure their stiffness. We insert those valves in an in vitro mock flow loop and measure the opening area of the aortic valve with high speed cameras as well as the instantaneous transvalvular flow. In this project(B), the aim is to define the function of the opening of the aortic valve with respect to flow, and thereby enable the determination of valve stiffness based on flow and opening of the valve only. This function would be used in the clincial practice to decide which patient suffering from low-flow, low-gradient aortic stenosis - a frequent diagnostic challenge for cardiologists in the clinics - should be offered an aortic valve replacement.

Prosthetic heart valve performance is investigated in compliant silicone aortic phantoms in an in vitro mock flow loop. In this project(C), we investigate the impact of aortic morphologies and compliance on valve kinematics and haemodynamic performance of prosthetic valves. We investigate and compare different patient-specific or standardised aortic morphologies and valve types and look at valve fluttering and contrast-enhanced blood flow in the aortic root.  

Additionally, a new prototype of mechanical valve featuring three leaflets is currently under evaluation in an international collaboration including company and university partners for which we are contributing to the in vitro evaluation of the valve performance [(A), (C)].

Current contributors: Lorenzo Ferrari (PhD student) (A)Dr Silje Ekroll Jahren (postdoctoral researcher) (C).

Former members: Dr Leonardo Pietrasanta, Dr Eric Buffle, Dr David Hasler and Dr Bernhard Vennemann.