Development and validation of a difficult laryngoscopy prediction model using machine learning of neck circumference and thyromental height

Table 2 Sensitivity (recall) and specificity and accuracy according to difficult laryngoscopy prediction model

	Threshold	Sensitivity (95CI)	Specificity (95CI)	Presision (95CI)	Accuracy (95CI)
BRF	0.54	0.90 (0.88–0.92)	0.58 (0.55–0.61)	0.23 (0.21–0.25)	60% (57–63)
XGB	0.23	0.54 (0.51–0.57)	0.78 (0.76–0.80)	0.25 (0.23–0.27)	69% (66–72)
LGBM	0.36	0.22 (0.20–0.24	0.91 (0.89–0.93)	0.28 (0.26–0.30)	83% (81–85)
MLP	0.37	0.49 (0.46–0.52)	0.60 (0.57–0.63)	0.19 (0.17–0.21)	61% (58–64)
LR	0.36	0.66 (0.63–0.69)	0.56 (0.53–0.59)	0.17 (0.15–0.19)	57% (54–60)

95CI 95% confidence interval, BRF balanced random forest, XGB extreme gradient boosting, LGBM light gradient boosting machines, MLP multilayer perceptron, LR logistic regression

ISSN: 1471-2253