Table 2.

Accuracy of the AI algorithm, dermatologists, dermatology residents, and general practitioners in overall conditions

	n	Top-1 Accuracy				Top-3 Accuracy
		AI algorithm	Average dermatologist	Average resident	Average general practitioner	AI algorithm
General	340	41.2%	60.1%	57.8%	49.3%	63.5%
Balanced	340	35.1%	45.2%	42.6%	33.1%	55.1%
Inflammatory	225	38.2%	63.3%	59.1%	50.8%	63.6%
Dermatitis	82	34.1%	61.4%	52.0%	48.8%	59.8%
Acne/rosacea	78	47.4%	81.2%	80.8%	73.9%	76.9%
Autoimmune	34	26.5%	37.3%	36.3%	16.7%	41.2%
Papulosquamous	17	52.9%	52.9%	47.1%	33.3%	82.4%
Infectious	52	50.0%	59.6%	58.3%	53.2%	69.2%
Viral	22	50.0%	65.2%	62.1%	51.5%	72.7%
Fungal	21	52.4%	68.3%	68.3%	71.4%	66.7%
Bacterial	8	50.0%	20.8%	25.0%	16.7%	62.5%
Parasitic	1	0%	66.7%	33.3%	0%	0%
Neoplastic	32	37.5%	50.0%	53.1%	41.7%	59.4%
Benign	28	42.9%	56.0%	59.5%	45.2%	64.3%
Malignant	4	0.0%	8.3%	8.3%	16.7%	25.0%
Alopecia	15	33.3%	44.4%	44.4%	31.1%	33.3%
Scarring	8	0%	25.0%	33.3%	16.7%	0%
Non-scarring	7	71.4%	66.7%	57.1%	47.6%	71.4%
Other	16	68.8%	52.1%	60.4%	47.9%	81.3%

AI, Artificial Intelligence; D, dermatologist; R, residents; GP, general practitioners.