Skip to main content

Table 5 Diagnosis performance of the three models

From: Preoperative prediction of extrathyroidal extension: radiomics signature based on multimodal ultrasound to papillary thyroid carcinoma

Group

Model

Sen

Sep

Acc

AUC (95% CI)

F1 score

 

Multimodal Model

54.5%

94.0%

87.0%

0.911 (0.866–0.957) *

0.691

Cross-validation

Clinical Model

12.1%

96.7%

81.5%

0.700 (0.593–0.807)

0.472

 

Radiomic Model

57.6%

94.7%

88.0%

0.908 (0.857–0.960) *

0.691

 

Multimodal Model

50.0%

92.6%

84.8%

0.716 (0.522–0.910)

0.571

Test set

Clinical Model

8.3%

96.3%

80.3%

0.676 (0.513–0.839)

0.421

 

Radiomic Model

58.3%

94.4%

87.9%

0.698(0.477–0.918)

0.636

  1. *Compared to AUC of Clinical Model in cross-validation cohort, there is a significant different (P < 0.05). Sen, sensitivity; Spe, specificity; Acc, accuracy; AUC, area under the curve; CI, confidence interval
Back to article page

BMC Medical Imaging

ISSN: 1471-2342

Contact us