HOUSTON, TX – A new artificial intelligence (AI) model has demonstrated the ability to accurately detect placenta accreta spectrum (PAS), a dangerous and often underdiagnosed pregnancy condition, before delivery, researchers announced today at the Society for Maternal-Fetal Medicine (SMFM) 2026 Pregnancy Meeting™. The model correctly identified all cases of PAS in a retrospective study, offering a potential breakthrough in reducing maternal morbidity and mortality associated with the condition.
PAS occurs when the placenta abnormally attaches to the uterine wall, a complication frequently linked to prior cesarean deliveries or other uterine surgeries. The incidence of PAS is rising in the United States, and its underdiagnosis can lead to life-threatening complications, including massive hemorrhage, organ failure, and death for the mother. Current screening methods, relying on risk factor assessment and ultrasound, can be inconclusive or result in misdiagnosis.
Researchers from Baylor College of Medicine in Houston retrospectively analyzed 2D obstetric ultrasound images from 113 patients identified as being at risk for PAS. These patients had given birth at Texas Children’s Hospital between 2018 and 2025, with a imply gestational age of approximately 30.89 weeks at the time of the ultrasound. The innovative AI program was then applied to these images.
The study revealed a high degree of accuracy. The AI model detected all instances of PAS present in the patient cohort. Researchers reported two false positive identifications, but crucially, no false negatives – meaning the model did not miss any actual cases of the condition. “Our team is extremely excited about the potential clinical implications of this model for accurate and timely diagnosis of PAS,” said Alexandra L. Hammerquist, MD, a maternal-fetal medicine fellow at Baylor College of Medicine. “We are hopeful that its use as a screening tool will help decrease PAS-related maternal morbidity and mortality.”
According to the Society for Maternal-Fetal Medicine, only about half of all PAS cases are diagnosed before delivery. This new AI model offers the potential to significantly improve that rate, allowing for better preparation and management of high-risk pregnancies. The research team has not yet announced plans for widespread clinical implementation or further trials.
