Table 3.

Summary of cohort-associated clinical metadata analysis to assess the model prediction accuracy.

Model variables	AUC	F score	Cutoff probability
PGE₂	0.843	0.778	0.317
Age	0.623	0.574	0.289
Estrogen	0.65	0.564	0.468
BMI	0.627	0.594	0.266
Post void residual (PVR)	0.56	0.504	0.299
pH	0.606	0.538	0.304
PGE₂, age, BMI, estrogen, PVR, and pH	0.882	0.761	0.308
PGE₂ and age	0.844	0.75	0.224
PGE₂, age, and estrogen	0.848	0.737	0.268
PGE₂, and estrogen	0.855	0.776	0.43
Age, BMI, estrogen, PVR, and pH	0.805	0.675	0.287

Leave-one-out cross-validation (LOOCV) procedure used to calculate the area under the ROC curve (AUC), F-score (predictive accuracy of the model), and the cutoff probability.