For all the miracles of modern medicine, the womb remains an enigma. A place of beginnings, yet often, a silent harbor for unseen battles. Congenital heart defects (CHDs)—malformations of the heart that develop before birth—are the most common birth anomalies worldwide, yet remain stubbornly elusive in prenatal screenings. The statistics are sobering: one in four babies born with CHD will require urgent medical intervention within their first year of life. And despite technological strides in prenatal care, the detection rate for these defects in routine ultrasounds remains worryingly low.
But what if artificial intelligence could change that?
At The Pregnancy Meeting™, the annual gathering of the Society for Maternal-Fetal Medicine, researchers presented a groundbreaking study that may very well redefine how clinicians spot the earliest signs of congenital heart defects.
AI as the Clinician’s Second Pair of Eyes
In a meticulously designed study, 14 obstetricians and maternal-fetal medicine subspecialists, with experience ranging from a single year to over three decades, analyzed 200 prenatal ultrasounds. Their task? To identify congenital heart defects—once without AI assistance and again with AI-based software acting as their silent collaborator.
The results were striking.
Regardless of experience level, the AI significantly improved their ability to detect cases that were suspicious for congenital heart defects. And it wasn’t just accuracy that improved—confidence levels soared, and diagnostic time reduced. A technology that doesn’t just make specialists more efficient, but more certain—surely, this is the kind of progress medicine has been waiting for.
But the findings also raise critical questions about access, expertise, and the role of AI in democratizing healthcare.
Bridging the Gap Between Specialists and Generalists
The uncomfortable truth is that in the United States, more than half of all prenatal ultrasounds are conducted by non-specialists—OB-GYNs and other medical professionals who may not have extensive training in fetal cardiology. The implications are clear: in a low-risk pregnancy, where a patient is unlikely to ever see a maternal-fetal medicine specialist, a critical heart defect could go unnoticed—a missed diagnosis that could mean the difference between a manageable condition and a fatal emergency.
“This is why the ability to detect congenital heart defects remains low, even in developed countries like the U.S.,”
explains Dr. Jennifer Lam-Rachlin, the study’s lead author and director of fetal echocardiography at Carnegie Imaging for Women in New York.
But with AI stepping into the equation, that gap may finally begin to close. “Our findings show that AI software significantly improved detection rates, not just among OB-GYNs but even among maternal-fetal medicine subspecialists. This has profound implications for neonatal outcomes and could truly change clinical practice.”
Which begs the question—should this be optional technology or standard practice?
A New Era in Fetal Medicine?
The AI-based software, developed by BrightHeart, is not merely a theoretical tool—it has already received FDA 510(k) clearance in November 2024, allowing its commercial use in real-world clinical settings.
Christophe Gardella, Ph.D., the chief technical officer at BrightHeart, underscores the mission behind the software:
“Most congenital heart defects happen in pregnancies that are considered low-risk. That means these cases are often in the hands of OB-GYNs, not fetal medicine specialists. We built this AI alongside experts to elevate detection rates, even among non-specialists, ensuring that diagnosis happens earlier and outcomes improve.”
It’s a bold promise, but the stakes could not be higher.
Every parent who walks into an ultrasound room carries hope and fear in equal measure. They watch a flickering heartbeat on the screen, trusting in the eyes of their doctor, trusting in the expertise behind the machine. But what if those human eyes are missing something?
For decades, prenatal detection of CHD has been plagued by inconsistency—an invisible lottery where some defects are caught, while others slip through undiagnosed. If AI can sharpen that vision, if it can ensure that more babies receive the early interventions they need, then this is no longer just an advancement—it’s a revolution.
A revolution in how we look at life before birth. A revolution in who gets a second chance at survival. And perhaps, most importantly, a revolution in how we balance human intuition with artificial intelligence, ensuring that medicine’s greatest gift—certainty—reaches every unborn child who needs it.
