Predictive placentas: Using artificial intelligence to protect mothers’ future pregnancies

New technology can decrease healthcare costs, allowing a majority of mothers and infants to have access to a microscopic placenta examination, according to a report in The American Journal of Pathology

Philadelphia, September 2, 2020

After a baby is born, doctors sometimes examine the placenta—the organ that links the mother to the baby—for features that indicate health risks in any future pregnancies. Unfortunately, this is a time-consuming process that must be performed by a specialist, so most placentas go unexamined after the birth. A team of researchers from Carnegie Mellon University (CMU) and the University of Pittsburgh Medical Center (UPMC) report the development of a machine learning approach to examine placenta slides in The American Journal of Pathology, published by Elsevier, so more women can be informed of their health risks.

One reason placentas are examined is to look for a type of blood vessel lesions called decidual vasculopathy (DV). These indicate the mother is at risk for preeclampsia—a complication that can be fatal to the mother and baby—in any future pregnancies. Once detected, preeclampsia can be treated, so there is considerable benefit from identifying at-risk mothers before symptoms appear. However, although there are hundreds of blood vessels in a single slide, only one diseased vessel is needed to indicate risk.

“Pathologists train for years to be able to find disease in these images, but there are so many pregnancies going through the hospital system that they don’t have time to inspect every placenta,” said Daniel Clymer, PhD, alumnus, Department of Mechanical Engineering, CMU, Pittsburgh, PA, USA. “Our algorithm helps pathologists know which images they should focus on by scanning an image, locating blood vessels, and finding patterns of the blood vessels that identify DV.”

Machine learning works by “training” the computer to recognize certain features in data files. In this case, the data file is an image of a thin slice of a placenta sample. Researchers show the computer various images and indicate whether the placenta is diseased or healthy. After sufficient training, the computer is able to identify diseased lesions on its own.

It is quite difficult for a computer to simply look at a large picture and classify it, so the team introduced a novel approach through which the computer follows a series of steps to make the task more manageable. First, the computer detects all blood vessels in an image. Each blood vessel can then be considered individually, creating smaller data packets for analysis. The computer will then access each blood vessel and determine if it should be deemed diseased or healthy. At this stage, the algorithm also considers features of the pregnancy, such as gestational age, birth weight, and any conditions the mother might have. If there are any diseased blood vessels, then the picture—and therefore the placenta—is marked as diseased. The UPMC team provided the de-identified placenta images for training the algorithm.

Example image and blood vessel patches from data set. A: A digitized whole slide image of a placental membrane roll [low magnification; haemotoxylin and eosin (H&E) stain]. To illustrate relative scale, the blue square indicates a single blood vessel. B: Image patch showing an example of a healthy blood vessel (high magnification; H&E stain). C: Image patch showing a decidual arteriole affected by early-stage decidual vasculopathy, characterized by smooth hypertrophic muscle around the blood vessel lumen (high magnification; H&E stain). Scale bars: 125 µm (B); 100 µm (C). Original magnification, x20 (A) (Credit: College of Engineering, Carnegie Mellon University).

“This algorithm isn’t going to replace a pathologist anytime soon,” Dr. Clymer explained. “The goal here is that this type of algorithm might be able to help speed up the process by flagging regions of the image where the pathologist should take a closer look.”

“This is a beautiful collaboration between engineering and medicine as each brings expertise to the table that, when combined, creates novel findings that can help so many individuals,” added lead investigators Jonathan Cagan, PhD, and Philip LeDuc, PhD, professors of mechanical engineering at CMU, Pittsburgh, PA, USA.

“As healthcare increasingly embraces the role of artificial intelligence, it is important that doctors partner early on with computer scientists and engineers so that we can design and develop the right tools for the job to positively impact patient outcomes,” noted co-author Liron Pantanowitz, MBBCh, formerly vice chair for pathology informatics at UPMC, Pittsburgh, PA, USA. “This partnership between CMU and UPMC is a perfect example of what can be accomplished when this happens.”


Notes for editors
The article is “Decidual Vasculopathy Identification in Whole Slide Images Using Multiresolution Hierarchical Convolutional Neural Networks,” by Daniel Clymer, Stefan Kostadinov, Janet Catov, Lauren Skvarca, Liron Pantanowitz, Jonathan Cagan, and Philip LeDuc ( ). It appears in advance of The American Journal of Pathology, volume 190, issue 10 (October 2020) published by Elsevier.

It is openly available at

Dr. Liron Pantanowitz is now Director of Anatomic Pathology at the University of Michigan Medical School.

This work was partially funded by the Center for Machine Learning and Health at Carnegie Mellon University through the Pittsburgh Health Data Alliance as well as the Office of Naval Research.

Full text of the study is also available to credentialed journalists upon request; contact Eileen Leahy at +1 732 238 3628 or Journalists wishing to interview the authors should contact Lisa Kulick, Carnegie Mellon University, at or Courtney Caprara, University of Pittsburgh Medical Center, at

About The American Journal of Pathology
The American Journal of Pathology, official journal of the American Society for Investigative Pathology, published by Elsevier, seeks high-quality original research reports, reviews, and commentaries related to the molecular and cellular basis of disease. The editors will consider basic, translational, and clinical investigations that directly address mechanisms of pathogenesis or provide a foundation for future mechanistic inquiries. Examples of such foundational investigations include data mining, identification of biomarkers, molecular pathology, and discovery research. High priority is given to studies of human disease and relevant experimental models using molecular, cellular, and organismal approaches.

About Elsevier
Elsevier is a global information analytics business that helps scientists and clinicians to find new answers, reshape human knowledge, and tackle the most urgent human crises. For 140 years, we have partnered with the research world to curate and verify scientific knowledge. Today, we’re committed to bringing that rigor to a new generation of platforms. Elsevier provides digital solutions and tools in the areas of strategic research management, R&D performance, clinical decision support, and professional education; including ScienceDirect, Scopus, SciVal, ClinicalKey and Sherpath. Elsevier publishes over 2,500 digitized journals, including The Lancet and Cell, 39,000 e-book titles and many iconic reference works, including Gray's Anatomy. Elsevier is part of RELX, a global provider of information-based analytics and decision tools for professional and business customers.

Media contacts
Eileen Leahy
+1 732 238 3628

Chhavi Chauhan, PhD, Director of Scientific Outreach
The American Journal of Pathology
+1 240 283 9724