AI analyzes placenta at birth to quickly identify problems in newborns and mothers

The tool, based on computer vision and AI, can help doctors quickly assess the condition of the placenta immediately after birth, leading to improved care for mothers and newborns.

Scientists at Northwestern Medicine and Penn State have developed a computer program called PlacentaVision that can analyze a routine photo of the placenta to detect abnormalities that may indicate infection and neonatal sepsis.

The researchers used cross-modal contrast learning, an artificial intelligence technology that can align and understand the relationships between different types of data, such as visual and textual. This helped create a program that can analyze images of placentas. To train the model, they collected an extensive dataset including placental images and pathology reports over 12 years, and examined the relationship between these images and health outcomes. Later, scientists built a model capable of making predictions based on new data.

According to the researchers, PlacentaVision was designed to be easy to use. The tool can even work through a mobile app or be integrated into medical record systems, allowing doctors to receive test results immediately after delivery.

The placenta plays a key role in maintaining the health of both mother and newborn, but doctors often do not examine the placenta after birth, especially in areas with limited medical resources. The scientists say using their tool can help prompt action, such as prescribing antibiotics for the mother or baby and closely monitoring the newborn for possible infection.