Every year, numerous mothers tragically die after childbirth due to heart-related complications, many of which could potentially be avoided. Researchers at Mayo Clinic suggest that early screening for heart conditions, particularly before pregnancy, could help identify women at risk and enable more effective care, ultimately improving maternal health outcomes.
Led by Dr. Anja Kinaszczuk and Dr. Demilade Adedinsewo, Mayo Clinic researchers tested artificial intelligence (AI) tools designed to detect undiagnosed heart issues in women of childbearing age. Using electrocardiogram (ECG) recordings and a digital stethoscope, the researchers aimed to find early signs of heart muscle weakness in women seen during routine primary care visits.
The findings, published in The Annals of Family Medicine, demonstrate that these AI tools have high diagnostic accuracy in identifying left ventricular ejection fraction (LVEF) below 50%, a key indicator of heart muscle weakness. The research involved two distinct groups of women, aged 18 to 49:
Group 1: 100 women scheduled for an echocardiogram, the gold standard for evaluating heart function, who also underwent standard clinical ECG and digital stethoscope recordings.
Group 2: 100 women attending routine primary care appointments, to gauge how often the AI tools would detect heart problems in a general population.
The AI-driven ECG tool showed impressive diagnostic performance with an area under the curve (AUC) of 0.94, while the AI-powered digital stethoscope, Eko DUO, outperformed it with an AUC of 0.98. These results highlight the strong accuracy of the tools in detecting heart issues. In Group 2, the AI-ECG detected heart problems in 1% of women, while the AI stethoscope identified issues in 3.2% of cases.
“Nearly half of all pregnancies in the U.S. are unplanned, and about 1% to 2% of women may have undiagnosed heart conditions. Our study suggests that these AI tools could help screen women before pregnancy, leading to better pregnancy planning, early treatment, and improved health outcomes,” said Dr. Adedinsewo, a cardiologist and senior author of the study.
This research builds on earlier studies, including a pilot study in the U.S. that evaluated AI tools for detecting pregnancy-related cardiomyopathy, as well as a randomized clinical trial in Nigeria involving pregnant and recently postpartum women. Together, these studies emphasize the growing potential of AI in transforming cardiovascular screening, especially for women of reproductive age.
Further studies are underway to explore how these AI technologies can be used to screen broader populations for heart muscle weakness. Mayo Clinic has licensed the technology to EKO Health for their digital stethoscope with embedded ECG electrodes and to Anumana for the 12-lead ECG. Some of the study’s authors and Mayo Clinic have a financial interest in these technologies, with any resulting revenue being reinvested into the clinic’s nonprofit mission of advancing patient care, education, and research.
You Might Be Interested In:
-
High Cannabis Use Associated with Increased Colon Cancer Mortality
-
Dopamine Circuit in the Brain Key to Unlearning Fear, MIT Study Finds
-
Montana Lawmakers Struggle to Increase Hospital Oversight Amid Heavy Lobbying
