Table 1 Summary of significant Mann–Whitney U tests for radiomic features created from pre-contrast images

From: Diagnostic value of radiomics and machine learning with dynamic contrast-enhanced magnetic resonance imaging for patients with atypical ductal hyperplasia in predicting malignant upgrade

Radiomic feature No-upgrade Upgrade p-value
Median (IQR) Median (IQR)
Skewness (FO)  − 0.322
(− 0.629 to 0.186)
 − 0.580
(− 1.013 to − 0.342)
0.011
Joint average (GLCM) 18.7
(16.2 to 20.3)
19.9
(17.2 to 22.0)
0.025
Sum average (GLCM) 37.3
(32.3 to 40.7)
39.8
(34.5 to 44.0)
0.025
Auto correlation (GLCM) 362
(286 to 434)
412
(313 to 495)
0.029
hglre (RLM) 358
(288 to 422)
397
(313 to 485)
0.036
lrhgle (RLM) 462
(377 to 581)
567
(427 to 659)
0.018
srhgle (RLM) 333
(266 to 392)
360
(294 to 437)
0.049
hglze (SZM) 347
(284 to 414)
389
(305 to 469)
0.042
lzhgle (SZM) 1027
(726 to 1462)
1281
(937 to 1965)
0.040
hgce (NGLDM) 363
(288 to 427)
399
(314 to 490)
0.038
hdhge (NGLDM) 1266
(936 to 1778)
1542
(1209 to 2280)
0.021
  1. Median values with associated interquartile range (IQR) are reported as non-parametric tests were performed. Features selected for advancement to model development are italicized
  2. FO first-order, GLCM gray-level co-occurrence matrix, RLM run-length matrix, SZM size zone matrix, NGLDM neighborhood gray-level dependence matrix, hglre high gray-level run emphasis, lrhgle long-run high gray-level emphasis, srhgle short-run high gray-level emphasis, hglze high gray-level zone emphasis; lzhgle large zone high gray-level emphasis, hgce high gray-level count emphasis, hdhge high dependence high gray-level emphasis