An artificial intelligence algorithm is able to detect with great precision the future occurrence of atrial fibrillation from a simple electrocardiogram, indicates a recent study carried out by researchers at the Montreal Heart Institute.
Atrial fibrillation (AF) is the most common form of arrhythmia, but it is often asymptomatic. Without treatment, it markedly increases the risk of stroke, cognitive decline, and dementia.
Analysis of the electrocardiogram with artificial intelligence would make it possible to predict the occurrence of AF beyond what a doctor is able to do by viewing the ECG with the naked eye. The use of this algorithm would therefore represent a major advance in the early detection of AF and the prevention of its complications.
The study also demonstrates that deep learning offers superior performance for predicting atrial fibrillation compared to traditional clinical predictors and genetic prediction.
The results of the study were published by theEuropean Heart Journal. They are based on the analysis of more than a million electrocardiograms from more than 250,000 patients. The findings were also presented on September 1 at the European Congress of Cardiology in London.
To see in video