CARLOS TALAYERO et al: SIMULATION OF BLOOD CLOT REMOVAL BY ASPIRATION THROMBECTOMY IN .. DOI 10.5013/IJSSST.a.19.05.18 18.1 ISSN: 1473-804x online, 1473-8031 print Simulation of Blood Clot Removal by Aspiration Thrombectomy in Cerebral Vessels using Geometry Optimization of the Aspiration Device Carlos Talayero, Gregorio Romero Department of Mechanical Engineering Universidad Politécnica de Madrid Madrid, Spain ctalayero@gmail.com; gregorio.romero@upm.es Gillian Pearce Department of Mechanical Engineering School of Engineering, University of Birmingham Birmingham, United Kingdom gpearce2011@gmail.com Julian Wong Department of Cardiac, Thoracic & Vascular Surgery National University Heart Centre Singapore Singapore julianwonguk@gmail.com Abstract — Thrombectomy by aspiration is one of the most effective systems for vessel recanalization. We present the results of a study on the modelling and elimination of blood clots in the arteries of the human body using Bond-Graph methodology. The modelling focuses on the clot and the distal end section of an aspiration device that improves the effectiveness of the treatment by reducing the risk of breaking the clot. The final model considers an elastic characterization of the blood clot and the possibility of achieving a process of progressive detachment of the clot from the vessel wall. An optimization process based on a design of experiments (DOE) is undertaken. The results show good agreement between the Bond-Graph techniques and the Finite Element Method models considered for validation (Computer Fluid Dynamics and nonlinear mechanics). Physical tests with gelatine also validate the results. We conclude that the proposed geometry will potentially improve the results of recanalization when blood clots are extracted from the arteries for a range of given parameters. Keywords - Blood clot removal; Aspiration Thrombectomy Device; Stroke; Bond-Graph. I. INTRODUCTION Stroke is a very common cause of death worldwide [1]. Strokes arise in the brain when the blood supply to a certain area of the brain is interrupted, for example, by occlusion involving a blood clot. The arteries that make up the Circle of Willis in the human brain are commonly affected by a stroke. In the last decade, several devices have been developed and used to deal with the removal of blood clots that arise during stroke [2]. Stentrievers, classified as mechanical thrombectomy devices (MTDs), are a recent class of stroke thrombectomy devices approved by the Food and Drug Administration (FDA) for the recanalization of occluded cerebral vessels in patients with acute ischemic stroke. However, these devices can carry potential risks such as breakage of moving parts, risk of penetration into the vessel wall and risks of embolization upstream due to thrombus fragmentation. That is one reason why they are sometimes combined with aspiration devices [3]. There is an aspiration extraction device called GPTAD (GP Thrombus Aspiration Device) that attempts to overcome some of these potential problems [4-5]. The device can potentially reduce the risk of embolization downstream because it does not have to make contact with the clot, during removal of the blood clot. In addition, it has no moving parts, and this potentially reduces the risk of breakage that may occur in the device. In vitro studies by Tennucci et al. [6] have also shown that it potentially reduces the risk of clot fragmentation. Preliminary studies [7-9] have dealt with the formation, composition and shape of blood clots attempting to find some general parameters that can reliably define their behaviour. The study we present in this paper however, is specifically focused on the suction and on the behaviour of the clot. The model has shown that the Bond-Graph technique is very useful in representing different simulation conditions, enabling the incorporation of different parameters in a very effective and direct way. In addition, it allows us to quickly obtain the boundary conditions for more complex calculation and modelling based on three- dimensional methods such as FEM. We note that although theoretical studies allow us to perform simulations under several varied conditions, such models need a validation process with more realistic data based, for example, on in- vitro and in-vivo testing rather than using parameters from the literature [3,6]. Additionally, real scale prototypes will improve the results of the physical testing. As the intention of the model is to determine, not only when the blood clot begins to move or break, but also the