Scientists at Northwestern Medicine and Penn State are developing a tool to train computer vision and artificial intelligence to evaluate placentas at birth for abnormalities.
A computer program, PlacentaVision, can analyze a photograph of a placenta after birth to detect signs of infection and neonatal sepsis, which can be life-threatening and affects millions of newborns globally, Northwestern Medicine said in a press release.
Related: Mayo Clinic, Phillips collaborate to speed up cardiac MRI with AI
The new tool is featured in a Dec. 13 study in the journal Patterns.
It's common to see placentas sent to the lab for testing, but an immediately available tool could be a game-changer, study co-author Dr. Jeffery Goldstein, director of perinatal pathology and an associate professor of pathology at Northwestern University Feinberg School of Medicine, said in the press release.
“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,” Goldstein said in the release.
Northwestern provided the largest set of images for the study, and Goldstein led the development and troubleshooting of the algorithms, the release said.
Penn State associate professor Alison Gernand, principal investigator on the project, conceived the original idea for the tool to serve her work around the globe, particularly among women who deliver in their homes due to a lack of health care resources, the release said.
Not a Modern Healthcare subscriber? Sign up today.
“Discarding the placenta without examination is a common but often overlooked problem,” Gernand said in the release. “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.”
PlacentaVision could be used in low-resource areas where pathology labs and specialists are not available to help doctors spot infections, Yimu Pan, a doctoral candidate in the informatics program at Penn State's College of Information Sciences & Technology and lead author on the study, said in the release.
“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,” Pan said.
Researchers developed a machine-learning model, PlacentaCLIP+, that provides for highly accurate analysis that can handle a range of real-world delivery conditions, including lighting, image quality and clinical settings, the release said.
Download Modern Healthcare’s app to stay informed when industry news breaks.
Teaching the model how to analyze a variety of samples required a large, diverse dataset of placental images and pathological reports spanning a 12-year period, it said.
“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,” Pan said. “The user-friendly app would allow doctors and nurses to photograph placentas and get immediate feedback and improve care.”
The researchers plan to make the tool even smarter, testing it in numerous clinical settings and including more types of placental features and adding clinical data to improve predictions, the release said.
This story first appeared in Crain's Chicago Business.