Study: Artificial Intelligence Enhances the Accuracy of Fetal Screening for Congenital Heart Defects

A new study has shown that using artificial intelligence programs can significantly improve fetal screening for the early detection of congenital heart defects.

The research relied on a tool developed by the medical company BrightHeart. The research team analyzed 200 ultrasound scans performed during the second trimester of pregnancy across 11 medical centers in two countries, including at least 100 scans with potentially concerning results.

Fourteen doctors participated in the study: seven obstetricians and seven specialists in high-risk pregnancies. They reviewed each scan in random order, either with or without AI assistance, to identify signs of serious cardiac abnormalities.

The results, published in the journal Obstetrics & Gynecology, revealed that physicians were able to detect a greater number of suspected defects in less time when using AI compared to traditional examinations.

The detection rate increased from 82% to more than 97%, the reading time decreased by 18%, and the confidence level in the results improved by 19%.

Commenting on the findings, Dr. Andrei Rebarber of the Icahn School of Medicine at Mount Sinai in the United States stated:

“Our study highlights the significant potential of AI-supported software to improve detection rates and reduce disparities in identifying congenital heart defects worldwide.”

He added:

“The future of prenatal diagnostic imaging looks promising with the use of AI software as a supportive tool for clinicians.”