Skip to main content
. 2021 Jul 27;11:705456. doi: 10.3389/fonc.2021.705456

Table 3.

The predictive performance of the models in each group.

Models Training cohort Validation cohort
AUC (95%CI) ACC (%) SEN (%) SPE(%) F score (%) Precision (%) Recall (%) AUC (95%CI) ACC(%) SEN(%) SPE(%) F score (%) Precision (%) Recall (%)
Type
ADC model 0.691 (0.590-0.783) 61.9 35.7 88.1 48.4 75.0 35.7 0.681 (0.615-0.741) 62.6 32.6 92.6 46.6 81.6 32.6
Radiomic model 0.839 (0.971-0.884) 75.8 65.3 82.3 72.9 82.7 65.3 0.867 (0.792-0.930) 76.2 76.2 76.2 76.2 76.2 76.2
Combined model 0.851 (0.804-0.895) 80.0 75.8 84.2 79.1 82.8 75.8 0.867 (0.798-0.931) 81.0 85.7 76.2 81.8 78.3 85.7
Grade
ADC model 0.817 (0.768-0.863) 73.2 78.9 67.4 74.6 70.8 78.9 0.762 (0.667-0.841) 69.0 81.0 57.1 72.3 81.0 65.4
Radiomic model 0.917 (0.884-0.948) 86.8 92.6 91.1 87.6 83.0 92.6 0.870 (0.800-0.932) 82.1 81.0 83.3 81.9 82.9 81.0
Combined model 0.959 (0.936-0.978) 90.5 91.6 89.5 90.6 89.7 91.6 0.880 (0.815-0.938) 82.1 83.3 81.0 82.4 81.4 83.3
DMI
ADC model 0.677 (0.616-0.736) 64.2 90.8 37.6 71.7 59.3 90.8 0.642 (0.548-0.732) 64.6 95.8 33.3 73.0 59.0 95.8
Radiomic model 0.828 (0.778-0.875) 77.1 65.0 89.0 74.0 85.5 65.1 0.751 (0.662-0.834) 67.7 54.2 81.2 62.7 74.3 54.2
Combined model 0.839 (0.792-0.883) 78.4 75.2 81.7 77.7 80.4 75.2 0.766 (0.685-0.847) 65.6 52.1 79.2 60.2 71.4 52.1
LVSI
ADC model 0.731 (0.677-0.783) 67.5 98.3 36.8 75.2 60.8 98.3 0.689 (0.605-0.777) 66.7 98.0 35.3 74.6 60.2 98.0
Radiomic model 0.774 (0.722-0.824) 75.2 71.8 78.6 74.3 77.1 71.8 0.668 (0.577-0.757) 65.7 58.8 72.5 63.2 68.2 58.8
Combined model 0.816 (0.769-0.86) 75.2 76.1 74.4 75.4 74.8 76.1 0.746 (0.662-0.827) 68.6 62.7 74.5 66.7 71.1 62.7
LNM
ADC model 0.657 (0.572-0.743) 66.1 98.3 33.9 74.8 60.0 99.3 0.638 (0.58-0.692) 66.5 99.3 33.8 74.4 59.8 98.3
Radiomic model 0.901 (0.867-0.933) 85.7 89.0 82.4 86.1 83.4 89.0 0.895 (0.844-0.942) 81.4 91.5 71.2 83.1 76.1 91.5
Combined model 0.910 (0.878-0.942) 84.9 97.8 72.1 86.6 77.8 97.8 0.897 (0.841-0.944) 81.4 93.2 69.5 83.3 75.3 93.2

AUC, area under the curve; CI, confidence interval; ACC, accuracy; SEN, sensitivity; SPE, specificity; ADC, apparent diffusion coefficient; DMI, deep myometrial infiltration; LVSI, lymphovascular space invasion; LNM, lymph node metastasis.