Artificial intelligence for differentiating COVID-19 from other viral pneumonias on CT: comparative analysis of different models based on quantitative and radiomic approaches

European Radiology Experimental

Table 1 Radiomic features extracted

		Number of features	Range (HU)		Bin width	Number of bins	Total number of features
RF1	First-order statistics	19	-1,020	180	5	240	19
RF2	GLCM	24	-1,020	180	5, 25, 50	240, 48, 24	72
RF2	GLSZM	16	-1,020	180	25, 100, 200	48, 12, 16	48

Number of first-order statistics (RF1) and GLCM and GLSZM (RF2) features extracted with PyRadiomics tool together with range of Hounsfield unit (HU), bin width, number of bins used, and the resulting total number of features extracted for each type. GLCM Grey-level cooccurrence matrix, GLSZM Gray-level size zone matrix, HU Hounsfield unit, RF1 First-order radiomic features, RF2 Second-order radiomic features