AI Tools Can Predict Left Ventricular Dysfunction in Women of Reproductive Age

WEDNESDAY, May 14, 2025 (HealthDay News) — Artificial intelligence (AI) tools are effective for detecting left ventricular systolic dysfunction (LVSD) among women of reproductive age seen in primary care, according to a study published online in the March/April issue of the Annals of Family Medicine.

Anja Kinaszczuk, D.O., from the Mayo Clinic in Jacksonville, Florida, and colleagues enrolled 100 women aged 18 to 49 years who had a primary care physician and a scheduled echocardiography at Mayo Clinic Florida (cohort 1) in a pilot cross-sectional study. On the date of echocardiography, 12-lead electrocardiography (ECG) and digital stethoscope recordings were performed. Probabilities for LVSD, defined as left ventricular ejection fraction <50 percent, were generated using deep learning for the 12-lead ECG and stethoscope (AI-ECG and AI-stethoscope) recordings. The prevalence of positive AI screening results when deployed for routine use was estimated in a second cohort of 100 participants seen in primary care (cohort 2).

In cohort 1, 5 percent had LVSD. The researchers found that the area under the curve was 0.94 and 0.98 for AI-ECG and AI-stethoscope (maximum prediction across all chest locations), respectively, for predicting LVSD. The prevalence of a positive AI screen was 1 percent for AI-ECG and 3.2 percent for AI-stethoscope in cohort 2.

“The use of AI-ECG and the AI-stethoscope appear to be effective in screening for LVSD among women of reproductive potential in a primary care clinic setting. These tools offer a rapid and cost-effective solution for preconception cardiovascular screening,” the authors write.

The Mayo Clinic has licensed the 12-lead electrocardiography technology to Eko Devices and to Anumana and may derive financial benefit from use of this technology.

Abstract/Full Text (subscription or payment may be required)