Unsupervised segmentation for inflammation detection in histopathology images

  • Authors:
  • Kristine A. Thomas;Matthew J. Sottile;Carolyn M. Salafia

  • Affiliations:
  • Department of Computer and Information Science, University of Oregon, Eugene, OR;Department of Computer and Information Science, University of Oregon, Eugene, OR;Placental Analytics, LLC, Larchmont, New York

  • Venue:
  • ICISP'10 Proceedings of the 4th international conference on Image and signal processing
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

Acute inflammation of the placenta is associated with an increased rate of perinatal morbidity and mortality. This paper presents a novel method for analyzing digitized images of hematoxylin and eosin (H&E) stained histology slides to detect and quantify inflammatory polymorphonuclear leukocytes (PMLs). The first step is isolating the tissue by removing the background and red blood cells via color thresholding. Next, an iterative threshold technique is used to differentiate tissue from nuclei. Nuclei are further segmented to distinguish between nuclei which match morphological characteristics of PMLs and those which do not, such as connective tissue fibroblasts as well as chorion and amnion epithelium. Quantification of the remaining nuclei, many of which are likely neutrophils, are associated with amniotic fluid proteomic markers of infection and inflammation, as well as histological grading of neutrophils in amnion, chorion and decidua, and in the umbilical cord.