Simulation of Blood Flow in Arterial Microvascular Anastomoses

The works described herein have augmented existing research by generating new knowledge on microvascular haemodynamics. Meaningful data has been produced and published, with the ultimate aim of contributing to reduction in free-flap failure rates secondary to anastomotic thrombosis. Computational Fluid Dynamics (CFD) and Finite Element Analysis (FEA) simulations have been combined with in vivo flow data to create increasingly realistic anastomotic models, verified using analytic and numeric techniques. Both hand-sutured and mechanically coupled vessels have been compared, with focus placed on suture positioning and its affect on local haemodynamics. As Shear Strain Rate (SSR) influences thrombus formation, areas of high SSR were targeted throughout. Computational simulations investigated approximations of anastomotic SSR for steady-state and transient flows, using Newtonian and non-Newtonian rheological models, with both compliant and rigid vessel walls. This body of research has generated several novel findings. Notably, it has developed and verified the first computational microarterial anastomosis simulation demonstrating realistic sutures, and incorporating a measured human arterial pulse. Of relevance to clinicians, it has demonstrated that the presence of sutures, and extremes of suture orientation and spacing, significantly influences approximations of anastomotic SSR. This may increase anastomotic thrombus risk. Coupled arterial anastomosis simulations were found to demonstrate more favourable haemodynamic properties, almost equivalent to those seen in pristine vessels. This provides some evidence, not only in favour of arterial coupling, but also for alternative non-suture methods of performing microarterial anastomoses. It has been demonstrated that steady-state CFD simulations underestimate SSR predictions when compared with a realistic pulsatile waveform. Investigating the influence of rheological models has shown that each non-Newtonian constitutive model predicts similar values of SSR at the vessel wall, as they approximate a Newtonian case. In addition, for compliant vessel walls, it has been demonstrated that the maximum principal strain in a sutured anastomosis is greater than that seen in both pristine vessels and coupled anastomoses. Finally, simulating sutured anastomoses using FEA in the way described herein is representative of the clinical picture immediately following surgery. Improvements in surgical techniques and patient outcomes are the ultimate goals of this work. Whilst this will take considerable further time and research to impact the surgical community as a whole, the author has already adapted his microsurgical technique based on the findings herein, and is disseminating these best practices directly to trainees. Importantly, the simulations performed and data gathered within the published work goes some way to fill pre-existing voids in the literature