An Algorithm for Arrhythmias
September 26, 2017 | Terry Sharrer
There is a high hurdle in applying artificial intelligence to clinical diagnostics—i.e. patients have to trust a machine to identify what’s wrong with them. In this piece, researchers at Stanford used a tiny, portable ECG device to collect 30,000 thirty-second recordings of heart arrhythmias and then used “deep learning” to identify the spectrum of heart beat disorders. The algorithms turned out to be as accurate as cardiologists reading the usual ECG graphs. Still, patient acceptance has yet to solidify. MORE
Image Credit: Stanford News and TechnologyReview.com