Mitral valve regurgitation is the unwanted backflow (leakage) of blood through the mitral valve in the left part of the human heart. Regurgitation may be due to a disease of the mitral valve itself (eg valve leaflet problem) or structural changes (dilatation) to the ventricle that prevent the mitral valve from performing its function properly.
Estimating the severity of regurgitation and understanding the underlying cause is an extremely complex task. Cardiologists usually use ultrasound images to estimate the problem based on the absolute leak volume, the properties of the jet through the valve, etc. These parameters are in turn (simultaneously) influenced by many other factors, including the condition and properties of the mitral valve, the differential pressure across the valve.
Recently, Prof. Frank Timmermans (UZGent) proposed a relatively simple useful parameter, the so-called "average pixel intensity" index that can be estimated from the intensity of pixels in a continuous wave Doppler recording, assessing the jet velocities. Previous studies have shown its added value over existing parameters, but further research is desirable for further refinement and a full understanding of the index.
The aim of thesis is to develop a numerical 3D multi-physics model to generate synthetic ultrasound data of the blood flow through a leaking mitral valve. Together with prof. Frank Timmermans, we will identify representative patient cases and use medical images to segment and reconstruct the 3D geometry of the left atrium, ventricle and the mitral valve in a closed position. The next step is to generate the CFD (computational fluid dynamics) model to simulate the regurgitant jet, whereby dynamic prolaps of (one of) the leaflet(s) is imposed using an overset meshing technique. Simulations will be performed in Ansys, Fluent. Given the high velocities, the flow is turbulent. The final step in the workflow is to couple the CFD results to Field II, software that allows to simulate the acoustic pressure field emitted by the ultrasound probe, and calculate the backscattered signal for image reconstruction. The image bellow shows an example of a synthetically generated Pulsed Doppler signal (taken from Van Canneyt et al, (2013), Journal of Vascular Access).
Once the model has been developed, it will be used to run a parameter study to assess the sensitivity and specificity of the newly proposed method and its absolute accuracy. Validation will be possible via experiments in a hydraulic bench model.