Comparison between physician assessment and artificial intelligence analysis in digital dermoscopy: evaluation of 71 dermoscopic cases

Artificial intelligence (AI)-based dermoscopic analysis is increasingly being used to assist dermatologists in the early detection and monitoring of pigmented skin lesions. However, discrepancies between AI-generated assessments and clinical evaluation may occur. This retrospective study aimed to evaluate the level of agreement between dermatologist assessment and AI-generated analysis in the dermoscopic evaluation of skin lesions. The study included 71 dermoscopic cases examined at two time points between September 2024 and January 2026 at a dermatologic clinic in Tirana, Albania. Dermoscopic images were acquired using a digital dermoscopy platform equipped with an integrated AI algorithm for automated lesion analysis. Agreement between the dermatologist and the AI system was observed in 58 cases (81.7%), whereas 13 cases (18.3%) demonstrated discordance. The 95% confidence interval for the agreement rate was 72.7%-90.7%, indicating a high level of concordance between clinical assessment and AI-based evaluation. These findings suggest that AI-assisted dermoscopic analysis demonstrates substantial agreement with clinical evaluation and may serve as a valuable supportive tool in dermatologic practice. Nevertheless, the observed discordant cases highlight the continuing importance of clinical expertise and established dermoscopic criteria in guiding patient management. AI systems should therefore be regarded as complementary decision-support tools rather than independent diagnostic systems.