Artificial intelligence (AI) enhances the detection of congenital heart disease (CHD) in prenatal ultrasounds, improving diagnostic accuracy and efficiency.
A recent study presented at The Pregnancy Meeting by the Society of Maternal-Fetal Medicine demonstrated AI’s effectiveness in detecting CHD during routine obstetric ultrasound examinations. The AI system analyzed two-dimensional ultrasound images and identified eight morphological markers associated with severe CHD, enabling timely referral for further assessment. This advancement in technology supports both obstetricians and maternal-fetal medicine specialists in diagnosing complex fetal conditions.
The research involved 200 ultrasound scans from 11 medical centers, with half of the scans containing potential signs of CHD. Fourteen physicians reviewed each scan both with and without AI assistance. The findings revealed that AI significantly enhanced detection rates, increasing the receiver operating characteristic (ROC) curve from 0.83 to 0.97, sensitivity from 0.78 to 0.94, and specificity from 0.76 to 0.97. Moreover, AI-assisted review reduced the average examination time from 274 to 226 seconds, demonstrating improved efficiency in clinical workflow.
According to researcher Lam-Rahlin, “Our results show that the AI-based software significantly improved the detection of ultrasounds that were suspicious for congenital heart disease, not only among obstetricians and gynecologists but also among maternal and fetal medicine specialists.” These findings highlight AI’s potential to revolutionize prenatal diagnostics by increasing accuracy, reducing diagnostic time, and supporting early intervention for critical fetal conditions.
The integration of AI in prenatal screening represents a significant step toward improving maternal and fetal healthcare. As AI technology continues to evolve, its application in obstetric imaging may further refine diagnostic capabilities, ultimately leading to better patient outcomes and optimized prenatal care.
