Assessment of artificial intelligence-aided chest computed tomography in diagnosis of chronic obstructive airway disease: an observational study

Egyptian Journal of Radiology and Nuclear Medicine

Table 7 Discriminating thresholds and diagnostic performance of D-value for distinguishing severe from moderate COPD using forced expiratory volume in the first second (FEV1) spirometric parameter as a reference (gold) standard

Coordinates of ROC curve	Threshold for achieving optimal sensitivity and specificity	Threshold for achieving near-90% sensitivity	Threshold for achieving near-90% specificity
AUC	0.705
Standard error	0.070
95%CI	0.569–0.842
p value	0.013*
Cut-off	≥ − 4.2	≥ − 4.3	≥ − 3.5
Sensitivity
TP/(TP + FN)	77.8%	88.9%	28.0%
Specificity
TN/(TN + FP)	60.0%	54.0%	89.7%

ROC curve = receiver operating characteristics curve, AUC area under ROC curve, CI confidence interval, TP true positive, FN false negative, TN true negative, FP false positive
* Significant p value < 0.05
** Highly significant p value < 0.01