Aussies Use CSIRO Supercomputer To Diagnose Sudden Death Heart Conditions

It's scary to think that people die from undiagnosed, almost impossible-to-detect cardiac problems all the time. The sad part is that some of these can be treated, if correctly diagnosed in a timely fashion. Thanks to half-a-decade of hard work from a team at Sydney's Victor Chang Cardiac Research Institute, and some help from a CSIRO supercomputer, we might be a step closer to understanding these mysterious afflictions.

The team, headed by Dr Adam Hill, group leader at the institute's Computational Cardiology Laboratory, made use of the formidable computing power at CSIRO's disposable to "simulate a couple of hundreds of hearts at once", according to a story by The Age's Rose Powell. The data gathered made it clear that electrocardiograms, or ECGs, contain "rich and untapped" information.

A better understanding of how the heart operates when a patient has different genetic conditions, could lead to more accurate diagnoses and treatments. The only issue at the moment is that it's all "theoretical"; we'll need even data and time to put the info to use:

Molecular cardiologist Chris Semsarian from the University of Sydney's Medical School told Fairfax Media the study was an exciting development for heart science. "The limitation of this study is that it is only a theoretical, computational model at this stage, but that is how many of our breakthroughs start," Dr Semsarian said. "More and more researchers are using computational models around the world, but to see it applied to Long QT is fairly unique, and great groundwork for further heart condition research."

The article mentions that the supercomputer used is "among the world's 10 most powerful", however, according to the most recent supercomputer Top 500 list, Magnus, the beefiest system housed at iVEC's Pawsey Centre, is a Cray XC40 with 35,712 cores. It comes in at #41, which is far outside the top 10.

I'm sure it's no slouch, though.

[The Age]

Image: Manel / Flickr, licensed under Creative Commons 2.0

Trending Stories Right Now