Advances in Clinical Imaging and Artificial Intelligence Integration for Dry Eye Diagnosis

Dry eye disease (DED) is a prevalent multifactorial chronic disorder of the ocular surface, characterized by a loss of tear film homeostasis and accompanied by ocular symptoms such as dryness, fatigue, and visual impairment. The complex pathophysiology of DED, involving inflammation and tissue damage, generates vast repositories of multimodal clinical data. These data offer opportunities for the application of deep learning methods, which have shown promise in detecting and quantifying subtle ocular abnormalities. This review systematically synthesizes recent advancements in deep learning-based DED assessment, categorized by critical clinical biomarkers: corneal fluorescein staining, meibography, tear meniscus height, tear film break-up time, and blink patterns. Furthermore, this review analyzes the current research trajectory and identifies future directions to establish reliable, interpretable, and generalizable deep learning frameworks for personalized DED diagnosis and long-term management.