Predicting the Grade of Prostate Cancer Based on a Biparametric MRI Radiomics Signature

Table 3

The process of feature selection using the selection step.

Lasso: cross validation	Spearman
“Min intensity”	“Min intensity”
“Histogram entropy”	“Histogram entropy”
“Correlation_AllDirection_offset1_SD”	“Correlation_AllDirection_offset1_SD”
“GLCMEntropy_AllDirection_offset1”	“GLCMEntropy_AllDirection_offset1”
“GLCMEntropy_AllDirection_offset4”	“GLCMEntropy_angle0_offset7”
“GLCMEntropy_AllDirection_offset7”	“Sum average”
“GLCMEntropy_angle0_offset1”	“HighGreyLevelRunEmphasis_AllDirection_offset4_SD”
“GLCMEntropy_angle0_offset4”	“Elongation”
“GLCMEntropy_angle0_offset7”
“GLCMEntropy_angle135_offset1”
“GLCMEntropy_angle135_offset4”
“GLCMEntropy_angle135_offset7”
“GLCMEntropy_angle45_offset1”
“GLCMEntropy_angle45_offset4”
“GLCMEntropy_angle90_offset1”
“GLCMEntropy_angle90_offset4”
“GLCMEntropy_angle90_offset7”
“Hara entroy”
“Sum average”
“Sum entropy”
“ShortRunHighGreyLevelEmphasis_AllDirection_offset1”
“ShortRunHighGreyLevelEmphasis_angle0_offset1”
“ShortRunHighGreyLevelEmphasis_angle0_offset7”
“ShortRunHighGreyLevelEmphasis_angle45_offset1”
“HighGreyLevelRunEmphasis_AllDirection_offset4_SD”
“Elongation”

LASSO, least absolute shrinkage selection operator.