Table 2.
Features used for the development of radiomics models obtained for each of the three different analyses (CE-T1-weighted, T2-weighted, CE-T1-weighted + T2-weighted)
| CE-T1-weighted | T2-weighted | CE-T1-weighted + T2-weighted | |||
|---|---|---|---|---|---|
| Feature | p-value | Feature | p-value | Feature | p-value |
| shape_Sphericity | 7.44E-07 | shape_Maximum2D DiameterColumn | 2.60E-07 | shape_Maximum2D DiameterColumn_T2w | 4.06E-07 |
| shape_Maximum2D DiameterColumn | 7.44E-07 | shape_Sphericity | 2.90E-07 | gldm_Small DependenceHighGray LevelEmphasis_T1w | 6.84E-06 |
| glrlm_GrayLevelNon Uniformity | 6.10E-06 | firstorder_Range | 2.90E-05 | firstorder_ TotalEnergy_T2w | 1.15E-04 |
| gldm_Smal lDependenceHighGrayLevelEmphasis | 6.72E-06 | firstorder_ TotalEnergy | 1.23E-04 | firstorder_ TotalEnergy_T1w | 3.64E-03 |
| firstorder_Range | 5.33E-03 | glrlm_RunPercentage | 7.52E-03 | firstorder_Range_T1w | 6.06E-03 |
| glszm_LargeAreaLow GrayLevelEmphasis | 1.92E-02 | ||||
The features used for the development of radiomics models belong to all radiomic feature families (morphological—shape, first-order statistical, texture—gldm, glszm, glrlm)
Adjusted statistically significant p-values derived from the Wilcoxon–Mann–Whitney test for the selected features are reported (p)