Skip to main content

Table 5 Results after application of the classifier to the test cohort showing lung nodule classification based on different combinations of CI- and iodine-derived texture features

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

 

Accuracy

F1/Dice

Sensitivity

Specificity

4 best CI-derived features

0.84

0.89

0.89

0.69

4 best CI-derived features + EntropyIM

0.86

0.91

0.91

0.71

All CI-derived features

0.87

0.91

0.92

0.71

All iodine-derived features

0.74

0.84

0.90

0.24

All CI-derived features + EntropyIM

0.87

0.92

0.93

0.67

All features

0.84

0.90

0.93

0.52