23 September 2016
Scientists are using powerful supercomputers to visualise how blood flows through implanted blood pumps, which prolong the lives of heart patients waiting for a donor. The work will help to reduce the number of prototypes needed for clinical testing – so saving on time, materials and costs – and potentially lead to the earlier availability of devices for patient use.
The goal of the research team, from the Science and Technology Facilities Council’s (STFC) Scientific Computing Department, is to assess if using Computational Fluid Dynamics in conjunction with High Performance Computing (HPC) can increase confidence in the use of computer models to design these complex medical devices.
Heart failures and diseases claim more than 17 million lives* across the world each year – more than all forms of cancer combined. The number is expected to grow to 23.6 million by 2030*. In the UK alone, 155,000 peopleⱡ died from heart diseases in 2014. It is estimated that seven million people in the UK are living with heart and circulatory diseases and healthcare costs could be as much as £11 billionⱡ.
The STFC Scientific Computing Department team answered a call from the US Food and Drug Administration (FDA) to assess whether computer models can accurately simulate the performance of blood pumps (known as Ventricular Assist Devices, or VADs).
Working with colleagues from FH Aachen University of Applied Sciences and FZ Jülich Research Centre in Germany, and EDF R&D, STFC’s Professor David Emerson and Dr Charles Moulinec developed simulations of how a centrifugal pump would behave with blood flowing through it at different rates, and using various rotation angles.
Professor Emerson explains, “The pump has a central rotor and blades which turn, helping the blood to flow. It would take several revolutions of the blades before it gets to a quasi-steady state where the blood flows smoothly, so we looked at results obtained at between five and 15 revolutions.”
The researchers used four million hours of computing time on the Blue Joule HPC facility at STFC’s Hartree Centre and the JuQueen HPC facility at FZ Jülich to perform the billions of calculations needed to produce faster, more accurate blood-flow simulations. One simulation involved 76 million elements, or computational cells, just to describe a pump’s dimensions.
Through these calculations they were able to predict where any damage to the blood might occur through turbulence in the flow, which could lead to blood clotting (thrombosis) or the breakdown of red blood cells (haemolysis). These are the two major life-threatening factors for patients depending on VADs.
“We also explored different velocities,” adds Professor Emerson. “This showed that tip vortices exist – giving the same sort of effect as water flowing quickly over rocks – when the blades turn at higher speeds. In a real device these vortices would make the pump shudder.”
The US FDA will collect the data from simulations carried out by all participating research teams and analyse the results in a ‘blind’ test – so the identity of the team will remain hidden and the data will be assessed on its own merit. The results will be compared with laboratory experiments carried out in parallel on the real devices and the conclusions are expected to be published later this year.
“Our results suggest that the design of these VADs need to be more blood-sensitive to reduce the risk of haemolysis and thrombosis,” says Professor Emerson. “Our work, and the work of other groups, can be used by the FDA to improve future devices, making them safer and available earlier for heart patients in the future.”
The team has also secured funding to develop the project further, which will involve assessing an existing mathematical model called ‘Large Eddy Simulation’ as a turbulence model for blood flow.
The research is funded by the Science and Technology Facilities Council.
Statistics: *American Heart Association; ⱡ British Heart Foundation
Last updated: 11 July 2019