A new technology has been developed that can predict cardiac arrhythmia 30 minutes before it occurs.
The research results, which developed a new technology that can predict atrial fibrillation (AFib) 30 minutes before it happens, were published in Patterns.
Cardiac arrhythmia refers to all abnormal heart rhythms and can take various forms. Generally, arrhythmia can cause discomfort in everyday life, but depending on its nature, the severity can vary, making it important to diagnose and treat the cause early.
AFib is a type of arrhythmia where the atria beat about 300-600 times per minute, preventing proper blood flow to the ventricles and inhibiting cardiac function. It is reported that approximately 59 million people worldwide suffered from AFib in 2019.
AFib does not threaten life in and of itself, but it increases the risk of complications such as stroke due to blood clots formed by AFib. Especially, recent research has shown that patients with AFib have a higher risk of heart attacks, heart failure, dementia, and gastrointestinal and liver diseases.
The researchers developed the AI model WARN (Warning of Atrial Fibrillation), which helps predict AFib for early diagnosis and intervention. Dr. Jorge Goncalves, the author of the research paper, described WARN as a deep learning model that produces an output of AFib prediction from a short input of 30 seconds of heart rate.
The researchers trained and tested the deep learning model using electrocardiogram data collected through Holter monitors worn by 350 people at a hospital in Wuhan, China.
The author emphasized the remarkable performance of AI in finding patterns when given a large data set. The researchers revealed that they can predict AFib 30 minutes before its onset with about 80% accuracy. The researchers added that such early detection allows medical professionals to proactively respond through prescriptions of antiarrhythmics and anticoagulants.
Most Commented