DCE-MRI Pharmacokinetic-Based Phenotyping of Invasive Ductal Carcinoma: A Radiomic Study for Prediction of Histological Outcomes

Table 2

Reduced feature set of each classification task. For each feature, the image from which it was extracted is indicated (if it is a first- or second-order feature), the feature name, and the value of the Mann-Whitney test. In bold are indicated the features that are significant, according to the Bonferroni correction for multiple comparison.

ER+/ER−	PR+/PR−	HER2+/HER2−	TN/NTN	Ki67+/Ki67−	lowGrade/highGrade
iAUC – GLCM Variance ( = 0.05)	– Std ( = 0.06)	– Median ( = 0.02)	– GLCM Variance ( = 0.03)	– Median ( = 0.0^−2)	– GLCM Entropy ( = 0.08⋅10^−1)
– GLCM Sum Average ( = 0.09)	TIRM – GLCM Entropy ( = 0.89)	postC – Skeweness ( = 0.05)	– Skeweness ( = 0.04)	– Energy ( = 0.02)	iAUC – GLCM Energy ( = 0.02)
postC – Uniformity ( = 0.11)	postC – Entropy ( = 0.09)	TIRM – Skeweness ( = 0.04)	– GLCM Energy ( = 0.05)	– GLCM Entropy ( = 0.03)	TIRM – GLCM Sum Average ( = 0.16)
TIRM – Mean ( = 0.20)	– GLCM Sum Average ( = 0.19)	– Skeweness ( = 0.09)	TIRM – Rms ( = 0.17)	– Mean ( = 0.00^−2)	– GLCM Variance ( = 0.07)
– GLCM Energy ( = 0.09)	TIRM – Mean ( = 0.22)	postC – Std ( = 0.10)	postC – Skeweness ( = 0.16)	– Energy ( = 0.00^−2)	TIRM – Mad ( = 0.10)
– GLCM Variance ( = 0.08)	iAUC – Kurtosis ( = 0.22)	– Skeweness ( = 0.08)	iAUC – Kurtosis ( = 0.10)	– Mean ( = 0.00^−2)	– Mad ( = 0.10)
– Skeweness ( = 0.13)	– GLCM Variance ( = 0.25)	– GLCM Entropy ( = 0.17)	– Kurtosis ( = 0.06)	– Rms ( = 0.00^−1)	iAUC – GLCM Entropy ( = 0.02)
TIRM – Kurtosis ( = 0.27)	– Variance ( = 0.06)	– Std ( = 0.15)	– GLCM Contrast ( = 0.09)	– Rms ( = 0.00^−2)	postC – Uniformity ( = 0.11)
postC – Entropy ( = 0.13)	postC – Uniformity ( = 0.09)	– Energy ( = 0.04)	– GLCM Dissimilarity ( = 0.12)	TIRM – GLCM Entropy ( = 0.075)	– GLCM Homogeneity ( = 0.05)
– Skeweness ( = 0.17)	– Mad ( = 0.07)	– Rms ( = 0.02)	– GLCM Variance ( = 0.12)	iAUC – Energy ( = 0.00^−2)	iAUC – Skeweness ( = 0.04)
– GLCM Entropy ( = 0.26)	– GLCM Correlation ( = 0.24)	– Mad ( = 0.03)	TIRM – Mean ( = 0.18)	postC – Skeweness ( = 0.12)	– GLCM Energy ( = 0.03)
– GLCM Entropy ( = 0.17)	– Skeweness ( = 0.19)	iAUC – Median ( = 0.04)	iAUC – GLCM Variance ( = 0.11)	– Median ( = 0.00^−2)	iAUC – Kurtosis ( = 0.04)
– GLCM AutoCorrelation ( = 0.13)	– Uniformity ( = 0.21)	iAUC – Mean ( = 0.04)	– GLCM Entropy ( = 0.19)	– GLCM Sum Average ( = 0.04)	iAUC – GLCM Variance ( = 0.08)
– GLCM AutoCorrelation ( = 0.21)	TIRM – Rms ( = 0.19)	– Max ( = 0.17)	– GLCM Autocorrelation ( = 0.23)	– Max ( = 0.00^−1)	– GLCM Dissimilarity ( = 0.07)
TIRM – Rms ( = 0.20)	– GLCM Correlation ( = 0.25)	postC – Max ( = 0.09)	postC – GLCM Entropy ( = 0.24)	– GLCM Autocorrelation ( = 0.04)	– GLCM Entropy ( = 0.15)
iAUC – Kurtosis ( = 0.29)	– GLCM Energy ( = 0.24)	postC – Variance ( = 0.10)	TIRM – Median ( = 0.20)	– Std ( = 0.00^−1)	iAUC – Mad ( = 0.14)
iAUC – GLCM Correlation ( = 0.28)	– Max ( = 0.10)	– Variance ( = 0.14)	TIRM – Kurtosis ( = 0.30)	– Variance ( = 0.00^−1)	postC – Skeweness ( = 0.22)
– Skeweness ( = 0.27)	iAUC – GLCM Autocorrelation ( = 0.23)	iAUC – Energy ( = 0.05)	TIRM – GLCM Energy ( = 0.23)	– GLCM Correlation ( = 0.05)	– GLCM Energy ( = 0.17)
TIRM – Median ( = 0.24)	– Kurtosis ( = 0.18)	postC – Median ( = 0.16)	TIRM – GLCM Variance ( = 0.27)	– Mad ( = 0.00^−1)	postC – Entropy ( = 0.17)
– GLCM Sum Average ( = 0.28)	TIRM – Std ( = 0.23)	– Std ( = 0.05)	– GLCM Sum Average ( = 0.32)	– Entropy ( = 0.00^−2)	iAUC – Std ( = 0.16)
postC – GLCM Entropy ( = 0.29)	TIRM – Variance ( = 0.23)	– Variance ( = 0.05)	– GLCM Homogeneity ( = 0.23)	postC – Kurtosis ( = 0.20)	TIRM – GLCM Autocorrelation ( = 0.16)
postC – GLCM Energy ( = 0.27)	– Energy ( = 0.20)	– Mean ( = 0.03)	– GLCM Entropy ( = 0.34)	Minimun diameter ( = 0.08)	TIRM – Max ( = 0.12)
– Kurtosis ( = 0.20)	– GLCM Variance ( = 0.25)	– GLCM Variance ( = 0.20)	– Mad ( = 0.35)	– Mad ( = 0.00^−1)	– Kurtosis ( = 0.06)
– Mad ( = 0.38)	postC – GLCM Energy ( = 0.27)	– Skeweness ( = 0.20)	– GLCM Homogeneity ( = 0.20)	– Energy ( = 0.00^−1)	– Skeweness ( = 0.12)
– GLCM Variance ( = 0.28)	– Entropy ( = 0.22)	postC – Rms ( = 0.18)	postC – GLCM Energy ( = 0.34)	– GLCM Correlation ( = 0.10)	– Skeweness ( = 0.22)