Zhanga R, Zhub L, Cai Z, et al. - A retrospective study was carried out in baseline ¹⁸F-flurodeoxyglucose PET/ CT (¹⁸F-FDG PET/CT) images of 135 individuals in order to investigate possible characteristics and develop classification models for differentiating benign and malignant lung lesions on the basis of CT-radiomics features and PET metabolic parameters extracted from PET/CT images. Via pairwise comparison in each fold, no important variations in receiver operating characteristic among models were noted. The possible peculiarities were discovered to be SurfaceVolumeRatio and SUVpeak. Hence, the classification models developed through CT-radiomics characteristics and PET metabolic parameters on the basis of PET/CT images have the substantial diagnostic capability on lung lesions.