. 2021 Oct 20;41(10):1569–1576. [Article in Chinese] doi: 10.12122/j.issn.1673-4254.2021.10.17

3.

训练集中不同模型效能比较的Delong检验

Delong test among NP, AP, VP and Combined models in the training set

Models	Training Set
Models	AUC (95%CI)	Sensitivity	Specificity	Delong
#: Delong test between NP and AP; ##: Delong test between NP and VP; ###: Delong test between AP and VP; : Delong test between AP and Combined model; : Delong test between VP and Combined model; **: Delong test between NP and Combined model.
NP	0.840 (0.778, 0.902)	69.7%	83.3%	0.665^#; 0.063^##
AP	0.819 (0.759, 0.877)	66.7%	82.5%	0.122^###; 0.328*
VP	0.730 (0.649, 0.803)	84.8%	50.8%	0.014**
Combined	0.861 (0.809, 0.910)	90.9%	69.2%	0.649***