State-Of-The-Art Methods for the Numerical Simulation of Aortic BMHVs

Abstract

Since the first clinical implantation of an artificial aortic valve by Dr. Charles A. Hufnagel in 1952 (Hufnagel et al., 1954), the use of such prostheses has gained strong interest and has become a routine treatment for severe heart valve failure. During the past 60 years, various mechanical heart valve designs have been developed for use in both aortic and mitral positions (Butany et al., 2003; Aslam et al., 2007). Nowadays, bileaflet mechanical heart valves (BMHVs) are widely preferred for aortic valve replacement because of their long lifespan. However, current BMHVs still induce pannus and thromboembolism, among other undesired side effects, which are believed to be due to non-physiological flow and turbulence generated by the valve leaflets (Sotiropoulous & Borazjani, 2009). One way to gain insight into the dynamics of a BMHV in order to improve its design is by experimental testing (Grigioni et al., 2004). Usually, in vitro testing is used, in which the functioning of the valve is assessed, for example, by using Doppler echocardiography (Dumont et al., 2002; Verdonck et al., 2002) or by visualizing the temporal and spatial flow field through velocimetry, like the laser Doppler anemometry (LDA) technique (Browne et al., 2000; Akutsu et al., 2001) or the particle image velocimetry (PIV) technique (Browne et al., 2000; Kaminsky et al., 2007). Also, the spectrum of the valve noise can be analyzed, as is done, for example, in Masson & Rieu (1998). Experimental in vivo testing is another option, using echocardiography and Doppler ultrasound to investigate the behavior of the valve after implantation in human patients (Bech-Hanssen, 2001; Aslam et al., 2007; Aljassim et al., 2008; Zogbi et al., 2009) or in animals (Yin et al., 2006). Numerical (“in silico”) methods can provide an alternative way to obtain relevant and detailed information for valve design optimization, since they are capable of solving the valve dynamics with a high degree of resolution in time and space (Kelly et al., 1999; Grigioni et al., 2004; Yoganathan et al., 2005; Dasi et al., 2009; Sotiropoulous & Borazjani, 2009). Moreover, they are considerably less time-consuming and less expensive during the research and development phase compared with experimental testing (Dasi et al., 2009) and are, therefore, particularly efficient for sensitivity studies (Verdonck, 2002). Unfortunately, the numerical simulation of a BMHV is a complex fluid-structure interaction (FSI) problem