One-dimensional computational hemodynamics modelling of healthy and diseased coronary arteries

Coronary artery disease (CAD) is one of the main causes of mortality in the developed countries. CAD is caused by the formation of atherosclerotic plaques in the arterial wall that progressively narrow the coronary vessel, potentially leading to severe events such as myocardial infarction. The study of hemodynamics is crucial to relate causes and effects of atherosclerosis, early detect and quantify stenoses, design stents and plan surgical interventions. Nowadays, computational fluid dynamics (CFD) models are widely used to investigate blood flow and pressure in healthy and diseased coronary arteries in addition to in vivo measurements and in vitro experiments. Since the solution of three-dimensional (3D) CFD models often requires excessive computational time, one-dimensional (1D) hemodynamic models have been increasingly applied, thanks to their several orders of magnitude lower computational complexity. In this context, the aim of the present thesis work is to computationally explore the advantages and limits of 1D models of healthy and diseased coronary arteries created with the open-source CFD software SimVascular by comparing the hemodynamic results with those obtained with corresponding 3D models. After a preliminary phase of tuning of the CFD model parameters, several 1D and 3D simulations were performed considering three coronary artery geometries reconstructed from computed tomography: a right coronary artery (RCA), a left circumflex coronary artery (LCX) and a left coronary artery (LCA) tree. Moreover, artificial stenoses were created in each vessel, studying three different degrees of occlusion: 0% (healthy), 50% (moderate) and 70% (severe) diameter reduction. For each geometry, rigid and deformable vessel wall simulations were performed under rest and hyperaemic conditions. The results showed a good agreement between the 1D and 3D models in healthy coronary arteries, in terms of computed pressure and flow waveforms. Conversely, in the case of the diseased vessels, the 1D models underestimated trans-stenotic pressure drop and consequently the fractional flow reserve (FFR) was overestimated. Indeed, hybrid reduced-order models (1D-0D), which include an algebraic (0D) pressure drop model accounting for stenosis, are needed to accurately estimate FFR. Anyway, the computational time required by 1D simulations was significantly reduced compared to 3D simulations, from several hours to few minutes. In conclusion, this thesis work pointed out the great potential of 1D computational hemodynamics modelling of SimVascular in reproducing realistic pressure wave propagation and flow distribution in healthy coronary arteries with short computational time, but also its structural limits in computing the pressure drop in the case of severe vessel narrowing.