Table 4.

Comparison of the accuracies of the model and 10 physicians for the prospective validation set.

	Model	Pediatric specialists	Other specialties	p-valuea (M vs P)	p-valuea (M vs O)
Task 1: normal vs. abnormal	0.8222	0.8489	0.8067	0.633	0.845
Task 2: crackle vs. wheezing	0.6774	0.7774	0.6613	0.128	0.922
Task 3: normal vs. crackle	0.6780	0.7864	0.7390	0.102	0.423
Task 4: normal vs. wheezing	0.8136	0.8441	0.7627	0.698	0.497

M model, P pediatric specialists, O physicians with other specialties.

^aChi-square test.