Stanford scientists develop algorithm to diagnose heart arrhythmias better than cardiologists

Researchers at Stanford have developed a computer algorithm that can scan hours of data collected by wearable electronics to identify instances of irregular heartbeats known as arrhythmias. The researchers developed a neural network based solution to identify fourteen type of arrhythmias. The algorithm was able to differentiate between dangerous kinds of irregular heartbeats and harmless irregularities. One of the most important applications of the algorithm is in remote areas where people may not have access to cardiologists. Awni Hannun, co-lead author of the paper says, “One of the big deals about this work, in my opinion, is not just that we do abnormality detection but that we do it with high accuracy across a large number of different types of abnormalities. This is definitely something that you won’t find to this level of accuracy anywhere else.”

The researchers compared diagnosis of the algorithm with six cardiologists who were independently asked to derive conclusions from the same data set used to test the algorithm. In most of the arrhythmias, the algorithm performed better than the cardiologists. Additionally, the algorithm can work around the clock and does not get tired. An immediate application of the technology is integration with fitness trackers that can send instant alerts to emergency services as soon as a dangerous and irregular heartbeat is detected.