Doctors at Mount Sinai's Raquel and Jaime Gilinski Department of Obstetrics, Gynecology and Reproductive Science have introduced an artificial intelligence (AI) tool to improve the accuracy of ultrasounds for detecting congenital heart defects in newborns. This marks the first large-scale implementation of such technology in New York City.
Congenital heart defects are among the most common birth abnormalities, with about 1 in 500 newborns affected by severe forms that require urgent intervention, according to data from the National Institutes of Health.
Carnegie Imaging for Women, which is affiliated with Mount Sinai and operates three locations in Manhattan, is the first center in New York City to use a Food and Drug Administration-approved AI software developed by BrightHeart. The software aims to make ultrasound readings more accurate and efficient.
A recent study published in Obstetrics & Gynecology, led by doctors from Mount Sinai West, found that using this AI technology improved detection rates of suspicious ultrasound findings for major congenital heart defects to over 97 percent. The study also reported an 18 percent reduction in reading time and a 19 percent increase in confidence scores among clinicians.
Researchers analyzed a dataset of 200 deidentified fetal ultrasound exams conducted between 18 and 24 weeks of gestation from 11 medical centers across two countries. Of these exams, half had at least one suspicious finding. Fourteen specialists—including obstetrician-gynecologists and maternal-fetal medicine experts—reviewed each exam both with and without AI assistance. They assessed whether each scan showed signs suspicious for congenital heart defects and rated their confidence.
The results indicated that AI-assisted interpretation led to better detection of lesions suggestive of severe congenital heart defects. The study concluded that integrating AI-based software into prenatal ultrasonography could enhance overall detection rates, clinician confidence, and efficiency.
"Our study should prompt and encourage future research into AI-assisted software's ability to improve detection rates, once integrated into clinical workflows, to reduce the variability and inequity of detection of congenital heart defects globally," said co-author Andrei Rebarber, MD, Director of the Division of Maternal-Fetal Medicine at Mount Sinai West and Clinical Professor at the Icahn School of Medicine at Mount Sinai. "The future for prenatal diagnostic imaging is bright when AI software is employed as an adjunct to physician interpretation."
