New AI-powered software designed to improve the quality of fetal ultrasound examinations has received authorization for clinical use from the U.S. Food and Drug Administration (FDA).

The solution, developed by U.S. startup BioticsAI, integrates with existing ultrasound systems and analyzes images in real time, detecting developmental abnormalities and alerting physicians to potential issues during prenatal ultrasounds, Euronews reports. Prenatal ultrasounds are widely used throughout pregnancy to identify possible problems in the developing fetus. These examinations can detect abnormalities affecting organs or body parts during key stages of formation.

When a fetal image is captured during an ultrasound, BioticsAI’s software analyzes it instantly. It assesses image quality, suggests adjustments to obtain a clearer view, and checks whether all parts of the baby’s body are visible, indicating what may be missing from the frame.

Drawing on patterns identified from global datasets, the system detects abnormalities and flags potential concerns, such as heart defects or limb malformations.

At the end of the examination, a report combining all findings is generated and made available to the physician after the ultrasound.

Developers say the software saves clinicians up to eight minutes per patient on documentation.

The FDA approved the software as a medical device for clinical practice, confirming that the product meets established performance standards and can be integrated into existing systems.

According to the latest data from the European Commission, serious congenital anomalies occur in 23.9 cases per 10,000 births across Europe.

A large-scale Cochrane review — conducted by the global independent organization that evaluates medical research — examined the diagnostic accuracy of standard prenatal ultrasounds, analyzing studies covering more than seven million pregnancies.

The review found that a single early ultrasound performed at 11–14 weeks detects only 38% of congenital anomalies. A mid-pregnancy scan, conducted at around 18–24 weeks, identifies 51% of abnormalities. When both scans are performed, the detection rate increases to 84%.

Similar developments are underway in Europe. French companies Diagnoly and Sonio Detect have also received approval to use AI in prenatal ultrasound examinations, with systems that automatically recognize fetal structures and detect cardiac abnormalities.