Heart failure is becoming one of the world’s fastest-growing health problems. The condition happens when the heart becomes too weak to pump enough blood to meet the body’s needs.
People with heart failure may feel tired, short of breath, weak, or unable to perform everyday activities. Over time, the condition can become severe and life-threatening.
Millions of people around the world live with heart failure, and the number continues to rise as populations age and diseases such as high blood pressure, diabetes, and obesity become more common.
In many low-income countries, the situation is especially serious. People often develop heart problems at younger ages and may have less access to hospitals, testing equipment, and specialist doctors. Because of these limitations, many patients are diagnosed too late, after significant heart damage has already occurred.
Now, researchers from UT Southwestern Medical Center and collaborators in Kenya have found that artificial intelligence may offer a low-cost way to detect early signs of heart failure before serious symptoms appear.
Their study, published in JAMA Cardiology, showed that a simple electrocardiogram, also known as an ECG, combined with an artificial intelligence algorithm, could accurately identify patients with hidden heart weakness in Kenya.
An ECG is one of the most common heart tests used worldwide. It measures the electrical activity of the heart through small sticky sensors placed on the skin. ECG machines are relatively inexpensive and widely available compared to more advanced heart imaging equipment.
The researchers focused on detecting a condition called left ventricular systolic dysfunction, often shortened to LVSD. This condition occurs when the left side of the heart cannot pump blood effectively.
LVSD is considered an important early warning sign for heart failure. If detected early, doctors may be able to start treatments that slow disease progression and improve long-term outcomes.
Normally, the best way to diagnose LVSD is through a heart ultrasound test called an echocardiogram. This test creates detailed images of the heart and allows doctors to measure how well it pumps blood.
However, echocardiograms are expensive and require trained specialists and advanced equipment. In many parts of Africa and other low-resource regions, access to these tests is limited.
To address this problem, the researchers tested whether AI-enhanced ECGs could act as a simpler screening tool.
The study involved nearly 6,000 patients receiving routine medical care at eight healthcare facilities in Kenya.
All participants received AI-assisted ECG testing. A smaller group of 1,444 participants also underwent echocardiograms so researchers could compare the AI results with the gold-standard heart imaging test.
The findings were very encouraging.
The AI algorithm identified LVSD in about 14% of participants who received echocardiograms.
One of the strongest results involved the test’s negative predictive value. The AI-ECG achieved a 99.1% negative predictive value, meaning that when the AI said a patient likely did not have LVSD, it was almost always correct.
This is especially important in resource-limited healthcare systems because it may help doctors confidently rule out heart problems in many patients without needing expensive imaging.
The AI system also showed very high sensitivity, correctly identifying more than 95% of people who actually had LVSD.
In addition, the AI demonstrated strong specificity, meaning it could also correctly identify many people who did not have the condition.
Researchers also found that positive AI results were linked to other dangerous heart changes, including thickening of the heart muscle and problems with the heart relaxing properly between beats.
These findings suggest the AI system may help identify broader heart disease risks, not just one specific problem.
The study was led by Dr. Ambarish Pandey from UT Southwestern Medical Center together with researchers from Kenya and other collaborators.
The researchers believe the technology could become an important screening tool in places where access to echocardiography is limited or unavailable.
Heart disease is increasing rapidly in sub-Saharan Africa, but healthcare systems often struggle with shortages of equipment, specialists, and funding. In many regions, people may need to travel long distances for advanced heart testing, and many hospitals cannot provide regular cardiac screening.
A low-cost AI-ECG system could help bridge this gap by allowing routine clinics to identify high-risk patients earlier and refer them for further care when needed.
The researchers say the findings show how artificial intelligence may help improve healthcare access globally, especially in underserved communities.
At the same time, scientists caution that AI tools should support doctors rather than replace them. Additional research will still be needed to understand how AI-ECG screening performs in other countries and healthcare settings.
Even so, the results represent an important step toward making heart disease detection faster, cheaper, and more widely available around the world.
The study highlights how combining simple medical tools with modern AI technology may eventually help prevent severe heart failure and save lives in places where advanced healthcare resources remain limited.
If you care about heart health, please read studies that yogurt may help lower the death risks in heart disease, and coconut sugar could help reduce artery stiffness.
For more information about health, please see recent studies that Vitamin D deficiency can increase heart disease risk, and results showing vitamin B6 linked to lower death risk in heart disease.
Source: UT Southwestern Medical Center.
