Table 4.
A performance summary of the radiomics models in the training and external validation cohorts.
| Accuracy | Sensitivity | Specificity | AUC | |
|---|---|---|---|---|
| Logistic regression | ||||
| Training cohort | 0.77 ± 0.08 | 0.61 ± 0.11 | 0.86 ± 0.10 | 0.84 ± 0.07 |
| External validation cohort | 0.75 | 0.65 | 0.84 | 0.85 |
| Random forest | ||||
| Training cohort | 0.82 ± 0.07 | 0.84 ± 0.10 | 0.73 ± 0.10 | 0.88 ± 0.06 |
| External validation cohort | 0.84 | 0.93 | 0.76 | 0.90 |
| Support vector machine | ||||
| Training cohort | 0.75 ± 0.07 | 0.52 ± 0.12 | 0.91 ± 0.08 | 0.81 ± 0.08 |
| External validation cohort | 0.71 | 0.74 | 0.68 | 0.80 |
Values of accuracy, sensitivity, specificity, and AUC of the three models in the training cohort are the average values after 30 holdout cross-validation, which were described as mean ± standard deviation (SD). AUC, areas under the curve.