Table 2. The performance of each kind of textural features.
Textural features | Sensitivity | Specificity | Youden | AUC | Accuracy | Precision | F |
Correlation | 0.69 | 0.55 | 0.24 | 0.63 | 0.65 | 0.82 | 0.75 |
Cluster tendency | 0.81 | 0.29 | 0.10 | 0.53* | 0.68 | 0.77 | 0.79 |
Difference-entropy | 0.63 | 0.52 | 0.15 | 0.58 | 0.60 | 0.80 | 0.70 |
Difference-mean | 0.67 | 0.48 | 0.15 | 0.57 | 0.63 | 0.79 | 0.73 |
Energy | 0.81 | 0.38 | 0.19 | 0.60 | 0.70 | 0.80 | 0.80 |
Entropy | 0.66 | 0.55 | 0.21 | 0.62 | 0.63 | 0.81 | 0.73 |
Homogeneity | 0.73 | 0.45 | 0.18 | 0.59 | 0.66 | 0.80 | 0.76 |
IDM | 0.66 | 0.54 | 0.19 | 0.60 | 0.63 | 0.81 | 0.73 |
Inertia | 0.67 | 0.57 | 0.24 | 0.63 | 0.65 | 0.82 | 0.74 |
Mean | 0.68 | 0.43 | 0.11 | 0.53* | 0.62 | 0.78 | 0.73 |
MP | 0.68 | 0.54 | 0.21 | 0.61 | 0.64 | 0.81 | 0.74 |
SD | 0.53 | 0.62 | 0.15 | 0.59 | 0.55 | 0.81 | 0.64 |
Sum-mean | 0.82 | 0.29 | 0.10 | 0.52* | 0.68 | 0.77 | 0.80 |
Sum-entropy | 0.58 | 0.62 | 0.19 | 0.63 | 0.59 | 0.82 | 0.68 |
Abbreviation used: AUC, the least area under the curve; IDM, Inverse difference moment; MP, Maximum probability; SD, Standard deviation; F, F_measure;
*P>0.05.