Skip to main content
Fig. 1 | Breast Cancer Research

Fig. 1

From: An ecological measure of immune-cancer colocalization as a prognostic factor for breast cancer

Fig. 1

Measuring spatial colocalization of immune and cancer cells through image analysis and spatial statistics. a Example H&E image of a breast tumor. Three sections obtained from different locations of the tumor were stained with H&E. b Automated image analysis was used to identify cell types (cancer, immune and stromal cells including fibroblasts and endothelial cells) in this image. c Density of cancer cells and immune cells per square after applying a square tessellation to this image; squares with less than a predefined amount of tissue would be excluded from analysis. d and e Schematic over an arbitrary spatial plane demonstrating how colocalization statistics can discriminate a highly segregated cell pattern from a highly colocalized cell pattern. f Significant correlation between the Morisita index and visual scoring for immune-cancer colocalization in 40 randomly selected samples (JT-test p = 0.0084); focal (disperse immune cell aggregates), low (mild infiltrate unassociated with cancer), moderate (some spatial association with cancer) and marked (dense infiltrate closely associated with cancer)

Back to article page