Thesis (Ph. D.)--University of New South Wales, 2005. / Also available online.

The aim of the research was to investigate the effects of three parameters of the artificial heart on the (+) dP/dt max. The study was conducted using a mock circulation which was connected to an artificial heart. The data were collected using the COMDU software developed for the computer which monitors the artificial heart. Stepwise regression analysis was utilized to test the three hypotheses. Two of the null hypotheses for the study could not be rejected, as the independent variable did not significantly affect (+) dP/dt max. Although the third hypothesis was accepted, the results were not clinically significant. Limitations in the study were multicollinearity among the independent variables, small sample size, and the inability of the mock circulation to represent human responses.

The Spiral Vortex (SV) ventricular assist device (VAD) was investigated by 2-component laser Doppler anemometry (LDA) while pumping a refractive index-matched blood analogue fluid. The VAD was operated under physiological conditions corresponding to 75% assist (4 litres/minute) or weaning from assist (2 litres/minute). Data were sampled on a 5-mm grid throughout most of the interior of the blood chamber, using two orthogonal LDA configurations from which 3D velocity data were synthesised. Data were subjected to statistical analysis of quasistatic time intervals and approximation by Fourier series. The velocity vector fields were explored statically (via 2D plots) and dynamically (using 3D animations of the reduced data). Reynolds stresses were computed and visualised in 2D. Fluid pathlines were simulated and plotted in 3D. The flow was found to be dominated by an irrotational vortex that accelerated and precessed in phase with the pumping diaphragm. Two unexpected flow structures, a rising, swirling near-wall layer in diastole and a reflection of the outflow vortex upon valve closure, enhanced washing of the walls. The thickness of the boundary layer was estimated to be 2 mm. Fluid velocities were generally lower than those reported in steady-flow studies on the SV VAD, although turbulence was comparable. Under the weaning mode, the coherence of the main vortex was degraded and flow recirculation was observed distal to the inflow port; this operating mode must be regarded as an indication for anticoagulation. In both pumping modes, turbulence was elevated in association with asymmetric buckling of the pneumatically driven diaphragm. Suboptimal orientation of the tilting-disc inlet valve gave rise to augmented turbulence production and skewing of the main vortex; similar results were obtained for an axisymmetric polymer (Jellyfish) valve, despite its advantageous haemodynamics. Flow stagnation was apparent where the inflow stream impinged on the wall, opposite the inflow port. The overall design of the SV VAD appears to almost ideal, in the context of current technology. However, elimination of recirculation/stagnation zones, especially in the weaning mode, remains a priority for the ultimate optimisation of haemocompatibility. Pulsatile VADs will probably never be entirely free of flow recirculation or stagnation, and published claims to the contrary probably reflect study limitations.

A non-invasive, in vitro method of estimating mean pulmonary artery pressure (PAP) was developed. This information was obtained by establishing a relationship between the pneumatic right drive pressure (RDP) and PAP waveforms. The RDP-PAP relationship was formalized into a series of multiple-linear regression equations for TAH cardiac cycles of known fill volume (FV). Correlation of computed estimates of PAP to actual measurements showed that these equations were greater than 92% accurate within 1.84 mmHg. In addition, while the RDP-PAP relationships were wholly dependent on FV, it was shown that they are independent of the manner in which FV was obtained. This method proved useful over the clinical operating range of the pneumatic heart driver, as well as over the normal physiological range of PAP in the human. Effectiveness of this method in vivo needs to be demonstrated.

The blood pumping chambers and drive pump for a totally implantable artificial heart have been developed at the biomedical engineering laboratory. The drive pump is driven by an electric motor and provides alternate pumping of the blood pumping chambers. The blood pumping chambers are diaphragm pumps and have uniquely captured valves. The pressure waveforms produced by the blood pumping chambers were measured at 26, 52 and 78 beats per minute and the hemolysis caused by the blood pumps was measured under pressures likely to be encountered in the body. The average hemolytic index over the three trials was 0.041 for the right blood pumping chamber and 0.040 for the left blood pumping chamber. / M.S.