Use of a Radiomics Model to Predict Tumor Invasiveness of Pulmonary Adenocarcinomas Appearing as Pulmonary Ground-Glass Nodules

Table 2

One-variate analysis for differentiating preinvasive lesions from IPAs.

Feature	P value
PM	0.0000446
Shape	0.05
Lobulated	0.001
Pleural	0.00003422
Age	0.003
FD	2.20E-16
FOS_entropy	0.001
FOS_ninetyfive	2.20E-16
GLCM_contrast1	0.00006972
GLCM_contrast2	2.83E-10
GLCM_dissimilarity	2.83E-10
GLCM_entropy	0.00009698
GLCM_homogeneity	2.19E-08
GLCM_idm	0.00002345
GLCM_intensity	9.10E-10
RLM_HIRE	0.004
RLM_HISRE	0.001
RLM_IV	0.001
RLM_LILRE	0.000001362
RLM_LISRE	0.0009
RLM_LRE	0.03
RLM_RLV	2.03E-14
RLM_RP	0.03
NGTD_Complexity	2.26E-07
NGTD_Contrast	0.00001651
normGLCM_contrast_1	0.0000161
normGLCM_dissimilarity	0.0000161
normGLCM_entropy	0.0001
normGLCM_intensity	0.02
NGLD_EN	0.001
NGLD_NNU	2.20E-16
NGLD_SM	0.0163
ISZ_HIZE	0.0004
ISZ_HISZE	0.001
ISZ_IV	3.16E-08
ISZ_LILZE	0.04
ISZ_LIZE	0.0014
ISZ_LISZE	0.001
ISZ_SZV	6.29E-16
ISZ_SZE	0.00005694

Note: IPAs=invasive pulmonary adenocarcinomas, PM=pure or mixed density, and FD=fractal dimension.