Table 2.
Univariate association between the selected radiomic features and prediction of malignancy.
| Characteristic | Overall, N = 255 1 | 0, N = 126 1 | 1, N = 129 1 | p-Value 2 |
|---|---|---|---|---|
| CONVENTIONAL_std | 23.4 (20.7, 26.7) | 24.1 (20.9, 27.1) | 22.9 (20.5, 25.9) | 0.10 |
| CONVENTIONAL_Skewness | 1.10 (0.74, 1.47) | 1.16 (0.84, 1.52) | 1.07 (0.64, 1.40) | 0.015 |
| CONVENTIONAL_Kurtosis | 4.53 (3.47, 5.93) | 4.84 (3.75, 6.46) | 4.00 (3.23, 5.25) | <0.001 |
| DISCRETIZED_ExcessKurtosis | 1.02 (0.26, 2.36) | 1.44 (0.49, 2.84) | 0.73 (0.02, 2.01) | <0.001 |
| GLCM_Contrast___Variance | 6 (4, 12) | 8 (5, 13) | 5 (3, 10) | <0.001 |
| NGLDM_Busyness | 1.53 (0.74, 2.92) | 1.30 (0.58, 2.50) | 1.83 (0.83, 3.75) | 0.019 |
| GLZLM_SZE | 0.61 (0.56, 0.67) | 0.62 (0.58, 0.68) | 0.58 (0.54, 0.66) | <0.001 |
1 Median (IQR). 2 Wilcoxon rank sum test.