Artificial Intelligence Could Enable Accurate, Inexpensive Screening for Atrial Fibrillation

August 3, 2019
Artificial intelligence (AI) can detect the signs of atrial fibrillation (AF) in an electrocardiograph (ECG), even if the heart is in normal rhythm at the time of a test, according to a study published in The Lancet. In other words, the AI-enabled ECG can detect recent AF that occurred without symptoms or that is impending, potentially improving treatment options. This research could improve the efficiency of the ECG. While common, AF is often challenging to diagnose. AF may not occur during a standard 10-second, 12-lead ECG, and people are often unaware of its presence. Prolonged monitoring methods, such as a loop recorder, require a procedure and are expensive. “When people come in with a stroke, we really want to know if they had AF in the days before the stroke, because it guides the treatment,” said Paul A. Friedman, MD, Mayo Clinic, Rochester, Minnesota. “Blood thinners are very effective for preventing another stroke in people with AF. But for those without AF, using blood thinners increases the risk of bleeding without substantial benefit. That is important knowledge.” For the current study, Zachi I. Attia, Mayo Clinic, and colleagues used approximately 450,000 ECGs to train AI to identify subtle differences in a normal ECG that would indicate changes in heart structure caused by AF. These changes are not detectable without the use of AI. Researchers then tested the AI on normal-rhythm ECGs from a group of 36,280 patients, of whom 3,051 were known to have AF. A single AI-enabled ECG identified AF with an area under the curve (AUC) of 0.87, sensitivity of 79%, specificity of 79.5%, F1 score of 39.2%, and overall accuracy of 79.4%. Including all ECGs acquired during the first month of each patient’s window of interest increased the AUC to 0.90, sensitivity to 82.3%, specificity to 83.4%, F1 score to 45.4%, and overall accuracy to 83.3%. “An ECG will always show the heart’s electrical activity at the time of the test, but this is like looking at the ocean and being able to tell that there were big waves yesterday,” concluded Dr. Friedman. “AI can provide powerful information about the invisible electrical signals that our bodies give off with each heartbeat -- signals that have been hidden in plain sight.” Reference: https://doi.org/10.1016/S0140-6736(19)31721-0 SOURCE: Mayo Clinic