Page 92 - 南京医科大学自然版
P. 92
第46卷第9期
·1362 · 南 京 医 科 大 学 学 报 2026年9月
0 10 20 30 40 50 60 70 80 90 100
Points
RSAD
0 0.5 1 1.5 2 2.5 3 3.5
3 5
PI⁃RADS score
2 4
2
Number of lesions
1 3
Suspicious for Yes
capsular invasion
No
Suspicious for No
lymph nodemetastasis
Yes
Total points
0 20 40 60 80 100 120 140 160 180
Risk
0.1 0.3 0.5 0.8 0.9 0.99
Points：standardized score；Total points：total score；Risk：risk of positive margin.
图2 前列腺癌术后切缘情况预测列线图
Figure 2 Nomogram for predicting postoperative surgical margin status in prostate cancer
A B
1.0 1.0
0.8 0.8
ORT：0.513（0.793，0.847） ORT：0.610（0.727，0.833）
Specificity 0.6 Specificity 0.6
0.4
0.4
AUC：0.897（0.874-0.921） AUC：0.841（0.793-0.888）
0.2 0.2
0 0
1.0 0.8 0.6 0.4 0.2 0 1.0 0.8 0.6 0.4 0.2 0
Specificity Specificity
ORT：optimal risk threshold.
图3 预测模型在训练集及（A）和验证集（B）中的ROC曲线
Figure 3 ROC curves of the predictive model in the training set（A）and the validation set（B）
A B
1.0 1.0
Observed probability 0.8 Observed probability 0.8
0.6
0.6
0.4
0.4
Bias⁃corrected
Bias⁃corrected
0.2 Apparent 0.2 Apparent
Ideal Ideal
0 0
1.0 0.8 0.6 0.4 0.2 0 1.0 0.8 0.6 0.4 0.2 0
Predicted probability Predicted probability
B=1 000 repetitions，boot Mean absoiute error=0.011（n=649） B=1 000 repetitions，boot Mean absoiute error=0.009（n=278）
Apparent：calibration curve of the original model；Bias⁃corrected：calibration curve after Bootstrap bias correction；Ideal：ideal calibration curve；
Mean absolute error；B=1000 repetitions，boot：Bootstrap resampling method with 1 000 repetitions.
图4 预测模型在训练集及（A）和验证集（B）中的校准曲线
Figure 4 Calibration curves of the predictive model in the training set（A）and the validation set（B）
