Construction of Artificial Intelligence System in Dermatology: Effectiveness and Consideration of Chinese Skin Image Database (CSID)

Abstract After more than 60 years of development, artificial intelligence (AI) has been widely used in various fields. Especially in recent years, with the development of deep learning, AI has made many remarkable achievements in the medical field. Dermatology, as a clinical discipline with morphology as its main feature, is particularly suitable for the development of AI. The rapid development of skin imaging technology has helped dermatologists to assist in the diagnosis of diseases and greatly improved the accuracy of diagnosis. Skin imaging data has natural big data attributes, which is important data for AI research. The establishment of the Chinese Skin Image Database (CSID) has solved many problems such as isolated data islands and inconsistent data quality. Based on the CSID, many pioneering achievements have been made in the research and development of AI-assisted decision-making software, the establishment of expert organizations, personnel training, scientific research, and so on. At present, there are still many problems with AI in the field of dermatology, such as clinical validation, medical device licensing, interdisciplinary, standard formulation, etc., which urgently need to be solved by joint efforts of all parties.