Communications of the ACM
AI-Based Method Can Identify Health Risks in Future Pregnancies
Image showing a decidual arteriole affected by early-stage decidual vasculopathy. Scale bar: 100 micron.
Credit: Carnegie Mellon University
A team of researchers from Carnegie Mellon University and the University of Pittsburgh Medical Center have developed a machine learning approach that automates the detection of abnormalities in digitized placenta slides, which could allow more women to be informed of risks. Examination of the placenta after a baby's birth can indicate health risks in future pregnancies, but the process is time-consuming and requires a specialist so most placentas go unexamined after a birth.
The researchers describe their work in "Decidual Vasculopathy Identification in Whole Slide Images Using Multiresolution Hierarchical Convolutional Neural Networks," published in The American Journal of Pathology, published by Elsevier.
One reason placentas are examined is to look for a type of blood vessel lesions called decidual vasculopathy. 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.
"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," says Daniel Clymer, Department of Mechanical Engineering at CMU. "The goal here is . . . to help speed up the process by flagging regions of the image where the pathologist should take a closer look."
From Elsevier
View Full Article
No entries found