Applications of the Non-invasive Skin Imaging Techniques and Image-based Artificial Intelligence in Rosacea: A Narrative Review.

BACKGROUND: Rosacea is a common chronic inflammatory dermatosis with complex pathophysiology and heterogeneous clinical manifestations. Despite its prevalence, no specific serological biomarkers exist for reliable diagnosis or disease monitoring. Current reliance on subjective clinical assessment underscores the need for objective and quantifiable evaluation methods. SUMMARY: This comprehensive review examines the current applications and research progress of non-invasive skin imaging modalities-including computer-aided imaging analyzers, dermoscopy, reflectance confocal microscopy (RCM), optical coherence tomography (OCT), high-frequency ultrasound (HFUS), and laser speckle contrast imaging (LSCI)-in rosacea management. We also discuss the emerging potential of image-based artificial intelligence (AI) for enhancing diagnostic accuracy and clinical decision-making. The integration of multimodal imaging with AI provides a more comprehensive and objective approach to rosacea management, enabling precise subtype classification, accurate severity assessment, and improved treatment monitoring. KEY MESSAGES: Multimodal non-invasive imaging combined with AI offers a more objective and comprehensive framework for rosacea diagnosis, subtype stratification, and treatment monitoring, supporting personalized management strategies. However, clinical adoption remains limited by insufficient evidence. Future efforts should focus on large-scale validation, standardization of imaging protocols, and development of AI models that integrate multimodal data to facilitate clinical decision-making.