Skip to main content

Table 2 Statistically significant univariate associations between radiomic features and Fibroblast cell type abundance

From: Predicting cell invasion in breast tumor microenvironment from radiological imaging phenotypes

Radiomic Feature Feature Type Correlation Coefficient Adjusted p Value1
Tumor mean pixel intensity in Pre-contrast Texture 0.42 0.011
Tumor mean pixel intensity in Post-contrast1 Texture/Kinetic 0.38 0.013
Tumor mean pixel intensity in Post-contrast2 Texture/Kinetic 0.39 0.011
Tumor mean pixel intensity in Post-contrast3 Texture/Kinetic 0.39 0.011
Mean intermediate WOS Kinetic 0.39 0.011
Mean quick WIS Kinetic 0.43 0.011
Mean intermediate WIS Kinetic 0.41 0.011
Mean slow WIS Kinetic 0.39 0.011
Mean quick PE Kinetic 0.41 0.011
Mean intermediate PE Kinetic 0.41 0.011
Mean slow PE Kinetic 0.39 0.011
Tumor contrast Texture 0.24 0.023
Tumor skewness in Post-contrast2 Texture/Kinetic −0.38 0.013
Tumor kurtosis in Post-contrast3 Texture/Kinetic −0.34 0.045
  1. 1 P values were adjusted for multiple comparisons using the Benjamini & Hochberg false discovery rate (FDR) method