Table 4. Predictive performance compared with plain radiography and bone scintigraphy.
LR High sensitivity Model | LR High specificity Model | SVM High sensitivity Model | SVM High specificity Model | Plain radiograph (Presence of vascular calcification) | Bone scintigraphy | |
---|---|---|---|---|---|---|
TP | 13 | 12 | 13 | 12 | 10 | 7 |
TN | 9 | 11 | 8 | 10 | 4 | 9 |
FN | 1 | 2 | 1 | 2 | 4 | 7 |
FP | 2 | 0 | 3 | 1 | 7 | 2 |
Sensitivity | 0.93 | 0.86 | 0.93 | 0.86 | 0.71 | 0.50 |
Specificity | 0.82 | 1 | 0.73 | 0.91 | 0.36 | 0.82 |
AUC | 0.95 | 0.92 | 0.54 | 0.66 | ||
AUC 95% CI | 0.86–1.0 | 0.80–1.0 | 0.31–0.71 | 0.42–0.88 |
LR, logistic regression; SVM, support vector machine; AUC, areas under the ROC curves; ROC, receiver operating characteristic; TP, true positive; TN, true negative; FN, false negative; FP, false positive.