Noninvasive prediction of node-positive breast cancer response to presurgical neoadjuvant chemotherapy therapy based on machine learning of axillary lymph node ultrasound

Table 4 The selected radiomic features

Variable	OR	[0.025	0.975]	S.E.	Z	P
lbp-2D_gldm_DependenceEntropy	0.573	0.382	0.860	1.230	0.068	0.003
lbp-2D_glszm_GrayLevelNonUniformityNormalized	1.411	0.987	2.020	1.201	6.593	0.026
original_shape2D_Elongation	1.258	0.926	1.709	1.170	4.332	0.062
log-sigma-0-2-mm-3D_glcm_Correlation	0.728	0.542	0.978	1.163	0.122	0.015
wavelet-LL_glszm_LargeAreaHighGrayLevelEmphasis	0.826	0.642	1.062	1.137	0.225	0.059
wavelet-LL_glcm_MCC	2.142	1.530	3.001	1.188	84.099	0.000
lbp-2D_glszm_SizeZoneNonUniformityNormalized	0.792	0.580	1.080	1.171	0.229	0.061
wavelet-LL_firstorder_Skewness	1.421	0.999	2.020	1.197	7.071	0.022
wavelet-LH_glcm_MCC	0.746	0.546	1.019	1.172	0.159	0.029

ISSN: 1479-5876