Table 3.

Comparison of the performance for a 5-classification condition by our artificial intelligence model and 4 human experts.

Test participants	Sensitivity	Specificity	Accuracy, %
Deep learning model	0.66	0.91	66.9
Laryngologist A (11 years of experience)	0.61	0.89	60.1
Laryngologist B (10 years of experience)	0.63	0.88	56.1
General ENTa C (8 years of experience)	0.54	0.88	51.4
General ENT D (14 years of experience)	0.42	0.85	43.2

^aENT: ear, nose, and throat.