Skip to main content

Table 1 Linear regression parameters (Pearson correlation coefficient (r), coefficient of determination (R2) and Spearman’s rank-order coefficient (ρ)) computed to evaluate the inter-observer variability according to the radiomics features family

From: MRI-based radiomics to predict lipomatous soft tissue tumors malignancy: a pilot study

 

Size

Shape

Intensity distribution

GLCM

GLRLM

GLZSM

NGTDM

Gabor filtering

r

0.89 ± 0.04 (0.84–0.93)

0.58 ± 0.20 (0.27–0.89)

0.90 ± 0.16 (0.49–0.99)

0.77 ± 0.08 (0.59–0.86)

0.76 ± 0.13 (0.57–0.95)

0.77 ± 0.17 (0.49–0.99)

0.82 ± 0.08 (0.73–0.93)

0.98 ± 0.02 (0.96–0.997)

R2

0.80 ± 0.08 (0.70–0.86)

0.38 ± 0.24 (0.07–0.79)

0.83 ± 0.24 (0.24–0.98)

0.6 ± 0.12 (0.35–0.74)

0.6 ± 0.2 (0.33–0.89)

0.62 ± 0.26 (0.24–0.99)

0.67 ± 0.13 (0.54–0.86)

0.96 ± 0.04 (0.92–0.99)

ρ

0.92 ± 0.01 (0.90–0.93)

0.62 ± 0.17 (0.46–0.93)

0.86 ± 0.06 (0.72–0.92)

0.77 ± 0.04 (0.68–0.81)

0.74 ± 0.06 (0.57–0.81)

0.74 ± 0.01 (0.53–0.84)

0.87 ± 0.06 (0.77–0.92)

0.88 ± 0.02 (0.85–0.91)

  1. Values reported are mean ± standard deviation (ranges)