Using time-series chest radiographs and laboratory data by machine learning for identifying pulmonary infection and colonization of Acinetobacter baumannii

Table 2 The performance of differences models in the classification of infection and colonization

Model	AUC	Accuracy	Sensitivity	Specificity	P value
Model 1	0.741(0.568,0.775)	0.587(0.432,0.730)	0.458(0.256,0.672)	0.727(0.498,0.893)	0.001
Model 2	0.720(0.551,0.777)	0.652(0.498,0.786)	0.667(0.447,0.844)	0.636(0.407,0.828)	0.004
Model 3	0.845(0.680,0.875)	0.783(0.636,0.891)	0.833(0.626,0.953)	0.727(0.498,0.893)	<0.001
Model 4	0.850(0.638,0.873)	0.761(0.612,0.874)	0.833(0.626,0.953)	0.682(0.451,0.861)	<0.001

Note: Data in brackets are 95% CI. Model 1 (baseline model), clinical baseline information + laboratory data and radiographic features of T₃; model 2, model 1 + the change value of between T₃ and T₁; model 3, model 1 + the change value of between T₃ and T₂; model 4 (multiple time-series model), model 2 + model 3. All statistical comparisons between the AUC values of models 1–3 were significant (p < 0.001). The best value(s) within each group are indicated with bold typeface

ISSN: 1465-993X