Aug. 2 (UPI) -- Researchers have developed a machine-learning method that can pinpoint a dangerous heart condition, new research shows.
With 90 percent accuracy, researchers were able to detect atrial fibrillation using artificial intelligence-guided EKG, according to a study published Friday in The Lancet. The method even worked when no symptoms were present in patients.
"When people come in with a stroke, we really want to know if they had atrial fibrillation in the days before the stroke, because it guides the treatment," Paul Friedman, chair of the at Mayo Clinic's Department of Cardiovascular Medicine and study senior author, said in a news release. "Blood thinners are very effective for preventing another stroke in people with atrial fibrillation. But for those without atrial fibrillation, using blood thinners increases the risk of bleeding without substantial benefit."
Normal 12-lead EKG tests last for only 10 seconds. But atrial fibrillation often comes and goes without people noticing it, making it notice with the test.
In many cases, patients with atrial fibrillation must be tested with a loop recorder, which involves prolonged monitoring to catch the condition when it recurs. But that method can be costly.
To develop a more efficient, less expensive solution, the researchers trained their artificial intelligence using 450,000 EKGs from the Mayo Clinic digital database. In the end, the AI method was able to pick up small structural changes in the heart that indicate atrial fibrillation.
The study included more than 36,000 patients, including 3,051 who were already diagnosed with atrial fibrillation before the study.
The researchers say this AI-guided EKG method is compatible with a smartphone or watch.
Atrial fibrillation, the most common form of heart arrhythmia, occurs when the heart beats too fast, too slowly or irregularly. The Centers for Disease Control and Prevention estimates between 2.7 and 6.1 million people in the United States have this condition.
Now, the researchers hope doctors can use this new AI-guided solution to prescribe the right treatment for patients with atrial fibrillation.
"An EKG will always show the heart's electrical activity at the time of the test, but this is like looking at the ocean now and being able to tell that there were big waves yesterday," Friedman said. "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."