Cannula position and Bernoulli effect contribute to leg malperfusion during extracorporeal life support with femoral arterial cannulation-an in silico simulation study†.

OBJECTIVES: Limb ischaemia during extracorporeal life support (ECLS) using femoral artery cannulation is frequently observed even in patients with regular vessel diameters and without peripheral arterial occlusive disease. We investigated underlying pathomechanisms using a virtual fluid-mechanical simulation of the human circulation. METHODS: A life-sized model of the human aorta and major vascular branches was virtualized using 3-dimensional segmentation software (Mimics, Materialise). Steady-state simulation of different grades of cardiac output (0-100%) was performed using Computational Fluid Dynamics (CFX, ANSYS). A straight cannula [virtualized 16 Fr (5.3 mm)] was inserted into the model via the left common femoral artery. The ECLS flow was varied between 1 and 5 l/min. The pressure boundary conditions at the arterial outlets were selected to demonstrate the downstream vascular system. Qualitative and quantitative analyses concerning flow velocity and direction were carried out in various regions of the model. RESULTS: During all simulated stages of reduced cardiac output and subsequently adapted ECLS support, retrograde blood flow originating from the ECLS cannula was observed from the cannulation site up to the aortic bifurcation. Analysis of pressure showed induction of zones of negative pressure close to the cannula tip, consistent with the Bernoulli principle. Depending on cannula position and ECLS flow rate, this resulted in negative flow from the ipsilateral superficial femoral artery or the contralateral internal iliac artery. The antegrade flow to the non-cannulated side was generally greater than that to the cannulated side. CONCLUSIONS: The cannula position and ECLS flow rate both influence lower limb perfusion during femoral ECLS. Therefore, efforts to optimize the cannula position and to avoid limb malperfusion, including placement of a distal perfusion cannula, should be undertaken in patients treated with ECLS.

Simulation of personalised haemodynamics by various mounting positions of a prosthetic valve using computational fluid dynamics.

Diseases of the cardiovascular system account for nearly 42% of all deaths in the European Union. In Germany, approximately 12,000 patients receive surgical replacement of the aortic valve due to heart valve disease alone each year. A three-dimensional (3D) numerical model based on patient-specific anatomy derived from four-dimensional (4D) magnetic resonance imaging (MRI) data was developed to investigate preoperatively the flow-induced impact of mounting positions of aortic prosthetic valves to select the best orientation for individual patients. Systematic steady-state analysis of blood flow for different rotational mounting positions of the valve is only possible using a virtual patient model. A maximum velocity of 1 m/s was used as an inlet boundary condition, because the opening angle of the valve is at its largest at this velocity. For a comparative serial examination, it is important to define the standardised general requirements to avoid impacts other than the rotated implantation of the prosthetic aortic valve. In this study, a uniform velocity profile at the inlet for the inflow of the aortic valve and the real aortic anatomy were chosen for all simulations. An iterative process, with the weighted parameters flow resistance (1), shear stress (2) and velocity (3), was necessary to determine the best rotated orientation. Blood flow was optimal at a 45° rotation from the standard implantation orientation, which will offer a supply to the coronary arteries.