Quantifying lung cancer heterogeneity using novel CT features: a cross-institute study

Insights into Imaging

Table 6 Intraclass coefficient of quantitative image features

Group	Feature	Between-scans	Between operators	Between-algorithms
1	Circularity	0.894	0.797	0.828
	Solidity	0.906	0.754	0.921
2	Variance	0.945	0.965	0.978
	P90	0.989	0.993	0.993
	Auto-correlation	0.971	0.957	0.938
	Sum-average	0.973	0.955	0.933
	Long-run emphasis mean	0.852	0.887	0.887
3	Kurtosis	0.964	0.974	0.990
	Mean	0.980	0.989	0.991
4	Energy	0.980	0.977	0.977
	A_skewness	0.972	0.988	0.993
5	Cluster-shade	0.857	0.696	0.574
6	Maximum-probability	0.962	0.958	0.951
	GLCM Energy	0.939	0.919	0.919
	GLCM Entropy	0.932	0.888	0.888
	GLCM sumEntropy	0.930	0.889	0.888
7	Long-run high gray-level emphasis mean	0.806	0.934	0.950
	Long-run high gray-level emphasis standard error	0.839	0.933	0.952
8	A_Long-run emphasis mean	0.852	0.827	0.907