1. A deep learning system for differential diagnosis of skin diseases
- Author
-
Kimberly Kanada, Vishakha Gupta, Vivek T. Natarajan, David H. Way, Susan Huang, Sara Gabriele, Clara H. Eng, Rainer Hofmann-Wellenhof, Dale R. Webster, R. Carter Dunn, Kang Lee, David Coz, Lily Peng, Jessica Gallegos, Greg S. Corrado, Peggy Bui, Guilherme de Oliveira Marinho, Nalini Singh, Dennis Ai, Yuan Liu, Ayush Jain, and Yun Liu
- Subjects
Male ,FOS: Computer and information sciences ,0301 basic medicine ,Native Hawaiian or Other Pacific Islander ,Skin Neoplasms ,Nurse practitioners ,Computer science ,Computer Vision and Pattern Recognition (cs.CV) ,Computer Science - Computer Vision and Pattern Recognition ,Eczema ,Economic shortage ,0302 clinical medicine ,Acne Vulgaris ,Photography ,Medical diagnosis ,Melanoma ,Image and Video Processing (eess.IV) ,Hispanic or Latino ,General Medicine ,Middle Aged ,Alaskan Natives ,Telemedicine ,030220 oncology & carcinogenesis ,Carcinoma, Squamous Cell ,Female ,Warts ,Adult ,Teledermatology ,medicine.medical_specialty ,Referral ,Primary care ,Skin Diseases ,Physicians, Primary Care ,White People ,General Biochemistry, Genetics and Molecular Biology ,Diagnosis, Differential ,03 medical and health sciences ,Deep Learning ,FOS: Electrical engineering, electronic engineering, information engineering ,medicine ,Humans ,Psoriasis ,Nurse Practitioners ,Medical physics ,Keratosis, Seborrheic ,Folliculitis ,Asian ,business.industry ,Deep learning ,technology, industry, and agriculture ,Electrical Engineering and Systems Science - Image and Video Processing ,Dermatitis, Seborrheic ,Black or African American ,030104 developmental biology ,Carcinoma, Basal Cell ,Indians, North American ,Artificial intelligence ,Differential diagnosis ,business ,Dermatologists - Abstract
Skin conditions affect an estimated 1.9 billion people worldwide. A shortage of dermatologists causes long wait times and leads patients to seek dermatologic care from general practitioners. However, the diagnostic accuracy of general practitioners has been reported to be only 0.24-0.70 (compared to 0.77-0.96 for dermatologists), resulting in referral errors, delays in care, and errors in diagnosis and treatment. In this paper, we developed a deep learning system (DLS) to provide a differential diagnosis of skin conditions for clinical cases (skin photographs and associated medical histories). The DLS distinguishes between 26 skin conditions that represent roughly 80% of the volume of skin conditions seen in primary care. The DLS was developed and validated using de-identified cases from a teledermatology practice serving 17 clinical sites via a temporal split: the first 14,021 cases for development and the last 3,756 cases for validation. On the validation set, where a panel of three board-certified dermatologists defined the reference standard for every case, the DLS achieved 0.71 and 0.93 top-1 and top-3 accuracies respectively. For a random subset of the validation set (n=963 cases), 18 clinicians reviewed the cases for comparison. On this subset, the DLS achieved a 0.67 top-1 accuracy, non-inferior to board-certified dermatologists (0.63, p
- Published
- 2020
- Full Text
- View/download PDF