AI has ability to detect future heart attack: study

A new study found that artificial intelligence could be used to help detect risk signs and possibly even prevent sudden cardiac death.

"When the data is fulsome and accurate and has a large enough sample size, AI will be able to identify patterns and correlations that humans might struggle to see, especially when they require two or more factors or have seemingly contrarian conclusions," Phil Siegel, the founder of the Center for Advanced Preparedness and Threat Response Simulation, told Fox News Digital.

Siegel's comments come after the results of preliminary research by the American Health Association found that AI was able to identify people who were at more than a 90% risk of sudden death, according to a report on the study in Medical Xpress.


According to the report, researchers analyzed medical information with AI by using registries and databases of 25,000 people from Paris and Seattle who had died from sudden cardiac arrest and 70,000 more people from the general population, matching the two groups by age, sex and residential area.

The AI then analyzed the data gathered with personalized health factors to identify people at "very high risk of sudden cardiac death." In addition, researchers created personalize risk equations for individuals by plugging in data for treatment of high blood pressure, history of heart disease and behavior disorders such as alcohol abuse.

Christopher Alexander, the chief analytics officer of Pioneer Development Group, also lauded AI's ability to cut through data to help in medical diagnostics, telling Fox News Digital that such tools are "useful for medical diagnosis or research because of its ability to do pattern recognition."


Doctor looks at heart with patient. (Credit: BSIP/UIG Via Getty Images)

"The AI can look at millions of disparate data points and find connections a human analyst would miss that could literally save lives. In fact, there is voice monitoring software that, when powered by AI, can note a tightening of the vocal cords that typically means a heart attack is imminent," Alexander said. "From detection to 911 dispatch, an ever-vigilant AI will never miss a warning sign and can seamlessly dispatch a better informed emergency medical team in that crucial first hour of medical care."

Meanwhile, Siegel expressed optimism about the use of AI for such applications, but he cautioned that developers will have to be careful to make sure "the sample isn’t biased or they don’t have enough data or some of the data is not accurate."

RELATED: New heart transplant method could expand donor hearts by 30%

"In those cases, they might reach incorrect conclusions or ones that only work for a small sample," Siegel said. "This is the challenge of using AI for good … the data and the models have to be complete, accurate and unbiased, or you might make things worse not better."


File: Heart health (Credit: Getty)

"These models can be the best use of AI for good but also can be misused. When done well, they will help doctors make earlier, better and more helpful diagnoses."

Find more updates on this story at