Performance of a deep neural network in teledermatology: a single‐centre prospective diagnostic study. (22nd November 2020)

• Record Type: Journal Article
• Authors: Muñoz‐López, C.; Ramírez‐Cornejo, C.; Marchetti, M.A.; Han, S. S.; Del Barrio‐Díaz, P.; Jaque, A.; Uribe, P.; Majerson, D.; Curi, M.; Del Puerto, C.; Reyes‐Baraona, F.; Meza‐Romero, R.; Parra‐Cares, J.; Araneda‐Ortega, P.; Guzmán, M.; Millán‐Apablaza, R.; Nuñez‐Mora, M.; Liopyris, K.; Vera‐Kellet, C.; Navarrete‐Dechent, C.
• Is Part Of: Journal of the European Academy of Dermatology and Venereology. Volume 35:Number 2(2021)
• Page Start: 546
• Page End: 553
• Publication Date: 2020-11-22
• Subjects: Dermatology -- Periodicals; Sexually transmitted diseases -- Periodicals; 616.5
• DOI: 10.1111/jdv.16979
• Languages: English
• ISSNs: 0926-9959

Abstract

Abstract: Background: The use of artificial intelligence (AI) algorithms for the diagnosis of skin diseases has shown promise in experimental settings but has not been yet tested in real‐life conditions. Objective: To assess the diagnostic performance and potential clinical utility of a 174‐multiclass AI algorithm in a real‐life telemedicine setting. Methods: Prospective, diagnostic accuracy study including consecutive patients who submitted images for teledermatology evaluation. The treating dermatologist chose a single image to upload to a web application during teleconsultation. A follow‐up reader study including nine healthcare providers (3 dermatologists, 3 dermatology residents and 3 general practitioners) was performed. Results: A total of 340 cases from 281 patients met study inclusion criteria. The mean (SD) age of patients was 33.7 (17.5) years; 63% ( n = 177) were female. Exposure to the AI algorithm results was considered useful in 11.8% of visits ( n = 40) and the teledermatologist correctly modified the real‐time diagnosis in 0.6% ( n = 2) of cases. The overall top‐1 accuracy of the algorithm (41.2%) was lower than that of the dermatologists (60.1%), residents (57.8%) and general practitioners (49.3%) (all comparisons P < 0.05, in the reader study). When the analysis was limited to the diagnoses on which the algorithm had been explicitly trained, the balanced top‐1 accuracy of the algorithm (47.6%) was comparable to the dermatologists (49.7%) and residents (47.7%) but superior to the general practitioners (39.7%; P = 0.049). Algorithm performance was associated with patient skin type and image quality. Conclusions: A 174‐disease class AI algorithm appears to be a promising tool in the triage and evaluation of lesions with patient‐taken photographs via telemedicine.