Heart failure during pregnancy is a serious and often overlooked condition. Its symptoms—such as shortness of breath, extreme fatigue, and difficulty breathing while lying down—can easily be mistaken for normal pregnancy discomforts.
However, new research presented at the European Society of Cardiology Congress reveals that an artificial intelligence (AI)-enabled digital stethoscope has doubled the detection rates of heart failure in pregnant women compared to traditional obstetric care. The full findings of the Mayo Clinic study are published in Nature Medicine.
The trial took place in Nigeria, where pregnancy-related heart failure is more prevalent than in any other part of the world.
The study showed that the AI-enhanced digital stethoscope was 12 times more likely than traditional methods to detect heart pump weakness at an ejection fraction below 45%. This level is significant as it indicates a specific type of heart failure known as peripartum cardiomyopathy.
“Early detection of this heart failure is crucial for the mother’s health and well-being,” said Dr. Demilade Adedinsewo, a cardiologist at Mayo Clinic and the study’s lead investigator.
“The symptoms of peripartum cardiomyopathy can worsen as pregnancy progresses or more commonly after childbirth, posing a significant risk to the mother’s life if her heart becomes too weak. Timely medical intervention can help manage the condition, but severe cases may require intensive care, a mechanical heart pump, or even a heart transplant if not managed with standard medical therapy.”
The randomized, controlled clinical trial involved nearly 1,200 participants, who were screened for heart conditions using either standard obstetric care or AI-enhanced tools. Mayo Clinic researchers had previously developed a 12-lead AI-electrocardiogram (ECG) algorithm designed to predict a weak heart pump, clinically known as low ejection fraction. This algorithm was further refined by Eko Health for use in its point-of-care digital stethoscope, which is approved by the U.S.
Food and Drug Administration (FDA) for detecting heart failure with low ejection fraction. The study found that doctors using the AI-enhanced stethoscope and ECG could identify weak heart function with high accuracy. Specifically, the digital stethoscope helped detect twice as many cases of low ejection fraction (below 50%), and doctors using it were 12 times more likely to identify an ejection fraction below 45% compared to usual care.
The AI-based tools were tested at three different ejection fraction levels, which are used in clinical diagnoses. An ejection fraction below 45% is the threshold for diagnosing peripartum cardiomyopathy. Below 40% indicates heart failure with reduced ejection fraction, which is treatable with specific medications that reduce symptoms and lower the risk of death. An ejection fraction below 35% signals severely reduced heart function, often requiring advanced heart failure therapies and possibly an implantable defibrillator if pump function does not recover. Patients in the intervention group underwent an echocardiogram at the start of the study to confirm the AI-generated predictions.
“This study shows that we can improve the detection of peripartum cardiomyopathy among women in Nigeria, but more research is needed,” Dr. Adedinsewo stated. “Our next steps include evaluating how easily Nigerian healthcare providers, including doctors and nurses, can adopt and use this tool and, more importantly, its impact on patient care. In the U.S., peripartum cardiomyopathy affects about 1 in 2,000 women, and as many as 1 in 700 African American women. Testing this AI tool in the U.S. will help assess its effectiveness across different populations and healthcare settings.”