A new tool using computer vision and artificial intelligence (AI) could make it easier for clinicians to quickly assess placentas after birth, improving care for both mothers and newborns. Developed by researchers from Northwestern Medicine and Penn State, the findings were recently published in the journal Patterns.
Image Credit: David A Litman/Shutterstock.com
The tool, called PlacentaVision, examines simple photos of placentas to identify abnormalities associated with infections and neonatal sepsis—a serious condition that impacts millions of newborns globally.
Placenta is one of the most common specimens that we see in the lab. When the neonatal intensive care unit is treating a sick kid, even a few minutes can make a difference in medical decision-making. With a diagnosis from these photographs, we can have an answer days earlier than we would in our normal process.
Dr. Jeffery Goldstein, Study Co-Author, Director, and Associate Professor Feinberg School of Medicine, Northwestern University
Northwestern contributed the largest set of images for the study, with Dr. Goldstein leading the development and refinement of the algorithms.
The idea for PlacentaVision came from Alison D. Gernand, an associate professor at Penn State’s College of Health and Human Development. Her experience in global health exposed her to the challenges faced in areas where placentas are often discarded without examination due to limited healthcare resources.
Discarding the placenta without examination is a common but often overlooked problem. It is a missed opportunity to identify concerns and provide early intervention that can reduce complications and improve outcomes for both the mother and the baby.
Alison D. Gernand, Associate Professor, Department of Nutritional Sciences, Penn State College
Why Early Examination of the Placenta Matters
The placenta plays a crucial role in the health of both mother and baby, yet it is rarely examined after birth, especially in resource-limited areas.
This research could save lives and improve health outcomes. It could make placental examination more accessible, benefitting research and care for future pregnancies, especially for mothers and babies at higher risk of complications.
Yimu Pan, Study Lead Author and Doctoral Candidate, College of Information Sciences and Technology
Early identification of placental infections using tools like PlacentaVision could allow clinicians to act quickly, the researchers explained. This might include administering antibiotics to the mother or baby and closely monitoring the newborn for any signs of infection.
The researchers emphasized that PlacentaVision is designed to be effective across a wide variety of healthcare settings.
Pan added, “In low-resource areas places where hospitals do not have pathology labs or specialists this tool could help doctors quickly spot issues like infections from a placenta. In well-equipped hospitals, the tool may eventually help doctors determine which placentas need further, detailed examination, making the process more efficient and ensuring the most important cases are prioritized.”
“Before such a tool can be deployed globally, core technical obstacles we faced were to make the model flexible enough to handle various diagnoses related to the placenta and to ensure that the tool can be robust enough to handle various delivery conditions, including variation in lighting conditions, imaging quality, and clinical settings. Our AI tool needs to maintain accuracy even when many training images come from a well-equipped urban hospital. Ensuring that PlacentaVision can handle a wide range of real-world conditions was essential,” stated James Z. Wang, Distinguished Professor and Study Principal Investigator at Penn State.
How the Tool Learned How to Analyze Pictures of Placentas
The researchers used a method called cross-modal contrastive learning, a type of AI that helps connect and understand relationships between different kinds of data—such as placental images and the text from pathology reports. This approach taught the program how to analyze photos of placentas. Over 12 years, the team collected a large and diverse set of images and reports, looking at how the images related to health outcomes. They also tested the system under different conditions, like varying lighting or photo quality, to make sure it would work reliably in real-world situations.
The result is PlacentaCLIP+, a powerful tool that can analyze placental photos with high accuracy to identify potential health risks. It’s been tested across different countries to ensure it performs well for a wide range of populations.
What makes PlacentaVision even more promising is its simplicity. The researchers aim to make it easy to use, possibly as a smartphone app or as part of medical record systems, so doctors can quickly get the information they need right after delivery.
Next Step: A User-Friendly App for Medical Staff
Pan stated, “Our next steps include developing a user-friendly mobile app that can be used by medical professionals with minimal training in clinics or hospitals with low resources. The user-friendly app would allow doctors and nurses to photograph placentas and get immediate feedback and improve care.”
The researchers are working to enhance the tool by incorporating additional placental features and integrating clinical data to make predictions even more accurate. They also plan to test the tool in a range of hospitals to ensure it performs reliably in different healthcare settings, contributing to both immediate care and long-term health research.
Gernand concluded, “This tool has the potential to transform how placentas are examined after birth, especially in parts of the world where these exams are rarely done. This innovation promises greater accessibility in both low- and high-resource settings. With further refinement, it has the potential to transform neonatal and maternal care by enabling early, personalized interventions that prevent severe health outcomes and improve the lives of mothers and infants worldwide.”
The research received support from the National Institutes of Health National Institute of Biomedical Imaging and Bioengineering (grant R01EB030130). The team also utilized supercomputing resources provided by the National Science Foundation’s Advanced Cyberinfrastructure Coordination Ecosystem: Services & Support (ACCESS) program.
Journal Reference:
Pan, Y., et al. (2024) Cross-modal contrastive learning for unified placenta analysis using photographs. Patterns. doi.org/10.1016/j.patter.2024.101097.