AI-Powered Stethoscope Helps Doctors Diagnose Twice as Many Cases of Pregnancy-Related Heart Failure

Heart failure during pregnancy is a dangerous and often unrecognized condition, as the usual symptoms – shortness of breath, extreme fatigue and breathing problems when lying down – are easily mistaken for typical pregnancy symptoms. Recent research presented at the European Society of Cardiology Congress as part of a Mayo Clinic study shows that a digital stethoscope equipped with artificial intelligence (AI) helped doctors detect twice as many cases of heart failure as a control group receiving usual obstetric care and examination. The full study results are published in Nature Medicine.

The study was conducted in Nigeria, where more women are affected by pregnancy-related heart failure than anywhere else in the world. The results also show that screening with the AI-enabled digital stethoscope is 12 times more likely to detect heart pump failure than conventional screening if the ejection fraction threshold is below 45%, which is the cut-off for a specific type of heart failure called peripartum cardiomyopathy.

How to Detect Heart Failure in Pregnancy with High Accuracy

“Early detection of this type of heart failure is important for the health and well-being of the mother,” says Dr. Demilade Adedinsewo, a cardiologist at the Mayo Clinic and leader of the study. The symptoms of peripartum cardiomyopathy can worsen as the pregnancy progresses or more often after delivery, endangering the mother’s life if her heart becomes too weak. Medication can help if the condition is recognized, but severe cases may require intensive care, a mechanical heart pump or sometimes a heart transplant if they cannot be controlled by drug therapy.

The randomized, controlled, open-label clinical trial involved nearly 1,200 participants who were screened for heart problems as part of usual obstetric care or with the help of AI. Mayo Clinic researchers had previously developed a basic 12-lead electrocardiogram (ECG) algorithm for predicting a weak heart pump, clinically known as a low ejection fraction. A version of this algorithm has been further developed by Eko Health for its digital point-of-care stethoscope, which is approved by the US Food and Drug Administration (FDA) to detect heart failure with low ejection fraction.

The researchers found that physicians performing AI-based screening with the digital stethoscope and a 12-lead ECG detected poor heart function with high accuracy. In the study cohort, the digital stethoscope detected twice as many cases with a low ejection fraction <50%, and physicians using the digital stethoscope were 12 times more likely to detect an ejection fraction <45% compared to usual care. The AI-assisted instruments were evaluated for three different levels of ejection fraction used in clinical diagnosis. Less than 45% is the cut-off for the diagnosis of peripartum cardiomyopathy. Less than 40% indicates heart failure with reduced ejection fraction and there is strong evidence that specific medication should be used to reduce symptoms and the risk of death. An ejection fraction of less than 35% indicates severely impaired cardiac pumping function, which often requires more intensive treatment, including advanced heart failure therapies and an implantable defibrillator if pumping function does not recover. All patients in the intervention group underwent an echocardiogram at the beginning of the study to confirm the predictions of AI.

According to Adedinsewo, this study provides evidence that researchers can better detect peripartum cardiomyopathy in women in Nigeria. However, there are still more questions to be answered. The next steps will be to assess the usability and acceptability of this tool by Nigerian healthcare providers (including doctors and nurses) and, most importantly, its impact on patient care. Peripartum cardiomyopathy affects approximately 1 in 2,000 women in the U.S. and as many as 1 in 700 African American women.