Texture analysis of iodine maps and conventional images for k-nearest neighbor classification of benign and metastatic lung nodules

Table 4 Results of the training cohort. KNN-based lung nodule classification with 5-fold cross validation using different combinations of CI- and iodine-derived texture features

	Accuracy	F1 / Dice	Sensitivity	Specificity
4 best CI-derived features	0.87 ± 0.03	0.91 ± 0.02	0.94 ± 0.03	0.69 ± 0.06
4 best CI-derived features + Entropy_IM	0.86 ± 0.04	0.90 ± 0.03	0.95 ± 0.04	0.65 ± 0.05
All CI-derived features	0.86 ± 0.02	0.91 ± 0.02	0.95 ± 0.02	0.64 ± 0.07
All iodine-derived features	0.73 ± 0.02	0.83 ± 0.02	0.92 ± 0.04	0.28 ± 0.08
All CI-derived features + Entropy_IM	0.86 ± 0.02	0.90 ± 0.02	0.96 ± 0.01	0.61 ± 0.08
All features	0.83 ± 0.06	0.89 ± 0.04	0.95 ± 0.03	0.56 ± 0.12

ISSN: 1470-7330