
BrightHeart presented promising data from two studies on its artificial intelligence (AI)-based software for screening congenital heart defects (CHDs) at the Society for Maternal-Fetal Medicine (SMFM) 2025 Pregnancy Meeting on January 30 in Aurora, CO.
In a statement, Paris-based BrightHeart noted that CHDs are the most common type of birth defect; however, up to 70% of CHD cases go undetected during routine prenatal ultrasounds due to the challenges of screening for them.
The software, which uses AI to analyze fetal heart ultrasound exams, was designed to close this screening gap, BrightHeart said. The presentation sessions at SMFM 2025 marked the software's formal introduction, underscoring its potential for reducing missed diagnoses with the study findings.
In one of the two studies, which employed a total of 877 ultrasound exams from 11 centers across multiple countries, the software provided conclusions for 98.8% of exams. Furthermore, among these conclusive exams, the software identified exams with signs associated with severe CHD with a sensitivity of 98.7% and confirmed the absence of such signs with a specificity of 97.7%.
Results from the other study, which used a dataset of 200 prenatal ultrasound exams reviewed by 14 physicians in the U.S. and European Union, demonstrated that BrightHeart's AI software significantly improved detection rates for CHD with a 15.3% increase in sensitivity while also improving efficiency with an 18% reduction in reading time.
BrightHeart's software received U.S. Food and Drug Administration (FDA) 510(k) clearance in November 2024.