The AI model uncovered unexpected combinations of these factors that might otherwise go unnoticed by clinicians…reports Asian Lite News
A new artificial intelligence-based model has identified previously unknown combinations of risk factors linked to serious pregnancy complications, including stillbirth. Developed by a team of researchers from the Universities of Utah and Brown, the model analyzed data from nearly 10,000 pregnancies in the U.S., incorporating social and physical characteristics such as blood pressure, medical history, fetal weight, and social support levels, along with the outcomes of each pregnancy.
The results, published in BMC Pregnancy and Childbirth, revealed that risk for stillbirth and other complications can vary up to tenfold for infants treated under the same clinical guidelines. Factors such as fetal sex, the presence of pre-existing diabetes, and fetal anomalies like heart defects were found to play a significant role in determining risk.
The AI model uncovered unexpected combinations of these factors that might otherwise go unnoticed by clinicians. Nathan Blue, a researcher from Utah’s Department of Obstetrics and Gynecology, noted that the findings could help advance personalized risk assessments and improve pregnancy care.
In a surprising twist, the model suggested that female babies could be at higher risk than males if the mother has pre-existing diabetes—an effect not typically observed in clinical settings, where female fetuses are usually considered to have slightly lower risks than males.
The team also focused on better risk estimation for fetuses in the lower 10 percent for weight. While current clinical guidelines call for intensive monitoring of these pregnancies, the researchers found that the risk for complications varies significantly within this group, challenging the one-size-fits-all approach to care. The AI model’s findings may lead to more nuanced and accurate pregnancy monitoring in the future.
