Melanoma is one of the most aggressive forms of skin cancer, and prognosis is highly dependent on early diagnosis. In recent years, artificial intelligence (AI), particularly deep learning systems applied to dermoscopic imaging, has emerged as a promising tool to improve early melanoma detection. This narrative review aims to analyze recent scientific evidence regarding the role of AI in early melanoma diagnosis, including its potential benefits and current limitations. A narrative literature review was conducted using PubMed, The Lancet Digital Health, JAMA Dermatology, Nature/npj Digital Medicine, Communications Medicine and European melanoma guidelines. Articles published between 2020 and 2026 related to artificial intelligence, dermoscopy and early skin cancer diagnosis were prioritized. The reviewed studies suggest that AI algorithms may achieve diagnostic performance comparable to dermatologists in selected clinical settings, especially through dermoscopic image analysis. However, significant limitations remain, including dataset bias, underrepresentation of diverse skin tones and the need for prospective validation in real-world clinical environments. Artificial intelligence represents a promising complementary tool in dermatology and venereology, although it should not currently replace clinical assessment or histopathological diagnosis.
Carlota Molina Padilla (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: