"A comparative analysis of transient and steady-state CFD simulations in coronary arteries "

The marked predisposition to develop atherosclerotic lesions, that narrow the coronary arteries affecting cardiac perfusion and which is associated risk of myocardial infarction make the coronary arteries an anatomical region of high clinical interest. The simultaneous presence of atherosclerotic lesions and disturbed flow in anatomical regions with more complex geometry such as arterial bifurcations has suggested that hemodynamic stresses play an important role in the pathogenesis of atherosclerosis. In this context, the computer aided engineering, coupled with medical imaging, which allows to perform patient-specific computational fluid dynamics (CFD) simulations is finding a wide application to identify possible link between local flow patterns and the development of atherosclerotic plaques in coronary arteries. Although transient CFD simulations are considered the best way to represent coronary hemodynamic, as they reproduce the pulsatile blood flow, their high computational cost, combined with the observation of low Womersley numbers and negligible values of oscillatory wall shear stress (WSS)-based descriptors, i.e. oscillatory shear index (OSI), found in literature, have led to reconsider the importance of steady-state CFD simulations in coronary arteries. Motivated from these considerations, the aim of this thesis was to conduct a preliminary CFD study on the impact that the use of steady-state instead of transient CFD simulations have on the hemodynamic descriptors of near-wall and intravascular flow, in angiographically healthy coronary arteries. Starting from the angiographic images of 16 patients provided by San Raffaele (Milan) hospital, the three-dimensional models of the vessels were reconstructed resulting in 13 left circumflex (LCX) and 3 left anterior descending (LAD) coronary arteries. Three-dimensional models were then discretized and steady-state and transient CFD simulations were performed assuming laminar flow. Blood was considered as an incompressible and homogeneous fluid and its non-Newtonian behaviour was modelled as a Carreau fluid. Regarding boundary conditions, the same inlet mean flow rate was assumed and prescribed as a parabolic velocity profile for both steady-state and transient CFD simulations. In particular, for transient CFD simulations, two different physiological waveforms, typical for LCX and LAD, measured using phase contrast magnetic resonance imaging (PC-MRI) were scaled according to mean flow rate used in steady-state CFD simulations. As for the outlets the portions of flow rate of each side branch were determined using a diameter-based scaling law. Arterial wall was assumed to be rigid and with no-slip conditions. The impact of pulsatility on near-wall hemodynamic was evaluated comparing the time averaged WSS (TAWSS) obtained from transient simulations with the steady-state derived WSS magnitude. Similar values were observed between the two approaches and low OSI values confirmed the poor oscillation of WSS vector during the cardiac cycle. Intravascular helical flow was evaluated comparing the helicity-based descriptors (h1, h2, h3, h4) and a more marked impact was observed. In conclusion, steady state CFD simulations are acceptable in the study of near-wall fluid dynamic while transient simulations are necessary to describe appropriately helical blood flow in healthy coronary arteries.