Table 2. Comparison of the LR model and ANN models.

Variable	LR models (mean ± std)	ANN models (mean ± std)
Sensitivity	99.5%±0.1%	83.1%±0.9%
Specificity	16.7%±2.3%	88.0%±2.9%
PPV	95.4%±0.4%	99.1%±0.2%
NPV	66.6%±6.6%	23.7%±1.9%
Accuracy	95.0%±0.4%	83.4%±0.8%
AUC	0.844±0.011	0.917±0.01

LR, logistic regression; ANN, artificial neural network; PPV, positive predictive value; NPV, negative predictive value; AUC, area under ROC (receiver operating curve).