Jennifer Hall, Ph.D., FAHA, American Heart Association Chief of the Institute for Precision Cardiovascular Medicine
copyright American Heart Association 

"I think that one of the easiest ways to describe it is that you're teaching a machine, hence, it's called machine learning or artificial intelligence, by teaching it with images in this case. So giving it electrocardiograms to review, and then you teach it by saying this is a normal electrocardiogram, and this is one that is not normal, and you let it see a lot of different electrocardiograms, millions, thousands, hundreds of thousands that are normal and hundreds of thousands that are not normal. And after a while, the machine, or the computer in this case, begins to understand pictures and images that look normal and those that do not look normal. And so, then over time, then the machine, or the computer in this case, can predict or understand which electrocardiograms are normal and which are not, and then we can use that to predict maybe who is at risk for a stroke or who is at risk of atrial fibrillation in the very near future."