Reproducibility of radiomic features in CT images of NSCLC patients: an integrative analysis on the impact of acquisition and reconstruction parameters

European Radiology Experimental

Table 2 FDR-adjusted p values for univariate and multivariable analysis for the effect of scanner and tube voltage

Features	Scanner (univar) FBP	Scanner (univar) IR60	Tube voltage (univar) FBP	Tube voltage (univar) IR60	Scanner (mixed)	Tube voltage (mixed)
shape_SurfaceArea	0.897	0.960	0.735	0.695	0.027	0.886
shape_VoxelVolume	0.936	0.960	0.784	0.695	0.190°	< 0.001°
Wavelet-glszm_SizeZoneNonUniformityNormalized*	0.264	0.905	0.005	0.016	0.996	0.005
Wavelet-glszm_SmallAreaEmphasis*	0.264	0.905	0.006	0.016	0.996	0.005
Wavelet-glcm1_Correlation*	0.462	0.905	0.144	0.130	0.561	0.018
Wavelet-glcm1_InverseVariance*	0.231	0.905	0.004	0.097	0.309	0.012

Only the features with significant FDR-adjusted p values at multivariate analysis
*HH filter
°In the model with VoxelVolume as the dependent variable, clinical volume was not used as independent predictor. FBP filtered backprojection, FDR false discovery rate, IR iterative reconstruction