A deep learning system for differential diagnosis of skin diseases

Nat Med. 2020 Jun;26(6):900-908. doi: 10.1038/s41591-020-0842-3. Epub 2020 May 18.

Abstract

Skin conditions affect 1.9 billion people. Because of a shortage of dermatologists, most cases are seen instead by general practitioners with lower diagnostic accuracy. We present a deep learning system (DLS) to provide a differential diagnosis of skin conditions using 16,114 de-identified cases (photographs and clinical data) from a teledermatology practice serving 17 sites. The DLS distinguishes between 26 common skin conditions, representing 80% of cases seen in primary care, while also providing a secondary prediction covering 419 skin conditions. On 963 validation cases, where a rotating panel of three board-certified dermatologists defined the reference standard, the DLS was non-inferior to six other dermatologists and superior to six primary care physicians (PCPs) and six nurse practitioners (NPs) (top-1 accuracy: 0.66 DLS, 0.63 dermatologists, 0.44 PCPs and 0.40 NPs). These results highlight the potential of the DLS to assist general practitioners in diagnosing skin conditions.

MeSH terms

  • Acne Vulgaris / diagnosis
  • Adult
  • Alaska Natives
  • Asian
  • Black or African American
  • Carcinoma, Basal Cell / diagnosis
  • Carcinoma, Squamous Cell / diagnosis
  • Deep Learning*
  • Dermatitis, Seborrheic / diagnosis
  • Dermatologists
  • Diagnosis, Differential*
  • Eczema / diagnosis
  • Female
  • Folliculitis / diagnosis
  • Hispanic or Latino
  • Humans
  • Indians, North American
  • Keratosis, Seborrheic / diagnosis
  • Male
  • Melanoma / diagnosis
  • Middle Aged
  • Native Hawaiian or Other Pacific Islander
  • Nurse Practitioners
  • Photography
  • Physicians, Primary Care
  • Psoriasis / diagnosis
  • Skin Diseases / diagnosis*
  • Skin Neoplasms / diagnosis
  • Telemedicine
  • Warts / diagnosis
  • White People