A team of researchers from the ARC Centre of Excellence for Mathematical and Statistical Frontiers (ACEMS) at QUT and their collaborators in the Computational Cardiovascular Science group at the University of Oxford have developed a novel method for constructing a virtual population for use in cardiac research.
- The Virtual population allows for drug experimentation that cannot be done on real people.
- Testing on virtual populations reduces the need for heart experimentation on animals.
- ACEMS researchers combined mathematical modelling with advanced statistical techniques to create this virtual population.
- The Virtual population was calibrated on data made up of heart cells from humans. A single cell was taken from about 350 people, of whom approximately half had atrial fibrillation.
ACEMS-Team-Group-Shot
Prof Kevin Burrage, Dr Pamela Burrage, Dr Brodie Lawson, Dr Chris Drovandi
Their research has just been published in Science Advances, and builds upon the concept of using virtual populations to simulate, for example, the effects of drug treatments on a certain population. The idea of creating a virtual, or ‘in silico,’ population, in itself isn’t new. It’s done by using mathematical models to best describe members of a certain population dataset. But the way it is done here offers novel insights into atrial fibrillation. Virtual populations allow researchers to not just look at what is happening to a group of people – but why it’s happening.
Virtual populations tackle a problem called variability. At the cellular level, no two cells are alike. At the tissue level, there’s a huge variation in the structure of how hearts are constructed from different cells. Yet all these different hearts still function.
The ACEMS researchers have come up with a calibration technique that really advances the accuracy of modelling an actual population. In other words, their virtual population is the best representation to date of an actual human data set. In this case, electrophysiological readings for individual heart cells taken from about 350 people.
“By doing these mathematical models and taking into account the variability, we can explore what sort of differences might lead to healthy versus unhealthy hearts and how these may differ across different groups,” said Dr Lawson.
The research in this study focuses on a specific problem where the heart beats irregularly, atrial fibrillation (AF). The normal rhythm of the heart is set off by tiny electrical signals that propagate from cell to cell, telling each cell when it needs to contract in order to produce a heartbeat. Atrial fibrillation occurs when this pattern of excitation becomes abnormal in the top chambers of the heart, the atria. This can cause a rapid and irregular heartbeat. It’s important to diagnose and treat AF because it can cause blood clots that can block the blood supply to your vital organs and lead to a stroke. It may also be a sign that you have heart disease.
“When you’re studying the heart and the electrophysiology of the heart, the differences between people are very significant and very important,” said Dr Lawson.
Pharmaceutical companies are starting to get on board to the idea of using these virtual populations. They can investigate new treatments, or even look back at why old treatments succeeded or failed.
“We can ask all kinds of questions about what if we apply a drug treatment to some group in our virtual population to see if it raises any red flags. A pharmaceutical company can’t do that test in any effective fashion,” said Prof Burrage.
Unlike previous attempts to construct virtual populations, Dr Lawson and Prof Burrage were able to draw on the statistical expertise available to them through ACEMS. In this case, they turned to Dr Chris Drovandi, an ACEMS Associate Investigator at QUT, who provided the statistical techniques which allowed the team to make smarter decisions on picking the models that would best fit the data and to Dr Pamela Burrage who provided deep insights into the nature of variability and Dr Nicole Cusimano who was previously an ACEMS Researcher at QUT and is now in BCAM, Bilbao, Spain.
“We have a very innovative way of calibrating the collection of models against the data we have,” said Prof Burrage.
Team members hope the new research cuts down on the need for animal testing.
“Yes, it’s a model, so some will ask, ‘do we trust it?’ That’s why we do our work and show, yes, you can. In another way, it’s more trustable because it’s for humans and it’s using humans, even if they’re virtual humans as opposed to animal testing which has all sorts of ethical questions and the questions surrounding the differences in the hearts,” said Prof Burrage.
Researchers also believe they will be able to use the calibration techniques they’ve developed to other problems outside of cardiac research.
Media contact: Tim Macuga, ACEMS Communications & Media Officer, 07 3138 6741, 0478 571 226, Timothy.Macuga@qut.edu.au
ABOUT ACEMS: We are an ARC Centre of Excellence for Maths and Stats with researchers at seven Australian universities: The University of Melbourne, Queensland University of Technology (QUT), University of Technology Sydney (UTS), The University of Adelaide, Monash University, the University of New South Wales (UNSW), and The University of Queensland. For more information, head to ACEMS.org.au Twitter: @ACEMathStats