Artificial Intelligence in Dermatology: A Scoping Review

This scoping review evaluates the integration of artificial intelligence (AI) in dermatology, focusing on its application in diagnosing skin conditions such as melanoma and alopecia areata. Analyzing 41 research articles from 2019 to 2025, the review highlights that AI, particularly deep learning neural networks, has achieved high diagnostic accuracy, with a median sensitivity of 84.06% and specificity of 97.84% for skin malignancies. However, challenges such as lack of external validation, potential conflicts of interest, and systematic bias due to unbalanced sample selection are noted. Despite these issues, AI shows promise in enhancing diagnostic accuracy, especially for less experienced practitioners, but further validation and methodological improvements are necessary.