An autonomous AI triage system deployed across two UK hospitals assessed 8,391 skin lesion patients over 16 months, with researchers estimating the specialist time saved could support roughly 8,500 additional face-to-face dermatology appointments. The findings were presented at the European Academy of Dermatology and Venereology (EADV) congress in Vienna, held September 30 to October 3, 2026. The system uses deep learning to autonomously discharge lower-risk patients while referring higher-risk cases to specialist clinicians; safety monitoring identified six false-negative discharges with no reported adverse outcomes.
