In vitro and in silico modelling of arteriovenous malformations for surgical planning

Abstract

Modelling vascular pathologies and associated interventions using computational fluid dynamics (CFD) techniques can improve pre-surgical planning. This study examines arteriovenous malformations (AVMs), which involve the direct shunting of high-pressure arterial blood into low-pressure venous circulation, leading to severe ischemic events (strokes, seizures, etc.) for which coil embolisation, aimed at complete flow blockage, is the preferred intervention technique. Vascular parameters, including blood rheology, patient-derived vasculature, and haemostasis, were computationally replicated to augment physiological accuracy, while the flow blockage of the embolising coil and clot formations were modelled through cell-based porous media approaches. This investigation evaluates AVM characteristics and treatment efficacy, emphasizing the impact of coil number and insertion angle on flow reduction and thrombogenesis. Results indicate that a minimum of three coils are required for successful embolisation, while the preferred insertion angle was determined to be that which is aligned with the blood vessel axis, i.e., a 0° angle of insertion.