Artificial Intelligence analysis of over half a million European and Chinese women reveals striking differences in the facial skin ageing process.

ABSTRACT

BACKGROUND: Artificial Intelligence (A.I) and deep learning-based algorithms are increasingly being used in dermatology following the emergence of powerful smartphones with high-resolution cameras. OBJECTIVES: To use an A.I-based algorithm, validated by dermatologists, to compare the evolution of the skin ageing process among Chinese and European women. METHODS: Selfie images were taken by 465 587 European and 79 016 Chinese women ranging from 18 to 85 and 18 to 69 years old, respectively, without facial skin diseases and who had access to a smartphone with a high-resolution camera (≥4 Megapixels). The selfies were analysed by facial skin diagnostic using a smartphone application to grade the severity of 9 facial signs (including wrinkles, sagging, vascular, pigmentation signs, pores). RESULTS: Wrinkles/texture, ptosis and sagging increased linearly with age in European women compared to lower scores and more gradual increase in the younger age-classes in Chinese women. In Chinese women, pigmentation signs increased regularly between 18 and 40 years, plateaued between 40 and 60 years, then increased in the over 60s compared to lower scores and a slower more regular increase with age in European women. Vascularization signs increased steadily with age in European women compared to no significant change in Chinese women. CONCLUSIONS: Marked differences were observed in the skin ageing process between European and Chinese populations, both in the prevalence of each facial ageing sign and their kinetics. Automatic grading performed on selfies and analysed by A.I is a fast and confidential method for quantifying signs of facial ageing and identifying the main issues for each population and age-class, which is of practical interest, as it will allow the development of tailored prevention and therapeutic measures.