In vivo validation of the in silico predicted pressure drop across an arteriovenous fistula.

Abstract

The creation of an arteriovenous fistula offers a unique example of vascular remodelling and adaption. Yet, the specific factors which elicit remodelling events which determine successful maturation or failure have not been unambiguously determined. Computational fluid dynamic (CFD) simulations are increasingly been employed to investigate the interaction between local hemodynamics and remodelling and can potentially be used to assist in clinical risk assessment of maturation or failure. However, these simulations are inextricably linked to their prescribed boundary conditions and are reliant on in vivo measurements of flow and pressure to ensure their validity. The study compares in vivo measurements of the pressure distribution across arteriovenous fistulae against a representative numerical model. The results of the study indicate relative agreement (error≈8-10%) between the in vivo and CFD prediction of the mean pressure drop across the AVFs. The large pressure drop across the AVFs coincided with a palpable thrill (perivascular vibration) in vivo and fluctuations were observed in the numerical pressure drop signal due to flow instabilities arising at the anastomosis. This study provides a benchmark of the pressure distribution within an AVF and validates that CFD solutions are capable of replicating the abnormal physiological flow conditions induced by fistula creation.