Montargil

Hydroxychloroquine HCQ is a widely prescribed drug for conditions like systemic lupus erythematosus and other autoimmune, dermatological, and oncological disorders. However, HCQ can cause retinal toxicity, which is challenging to detect early and can worsen even after stopping the drug. This study aims to develop an automated method using deep learning technology to improve early detection and prediction of HCQ-related eye damage, potentially reducing the burden on clinicians. The study involves collecting images typically used in HCQ toxicity screening, such as fundus photography and optical coherence tomography OCT with autofluorescence. Participants are grouped into those with HCQ intake who have retinopathy and those without retinopathy. Both groups will undergo OCT scans analyzed by the RETINAI deep learning algorithm to assess retinal changes and predict toxicity development. Participants will provide images that are assessed by three blinded readers to validate findings. The study will measure the success of the automated screening method over one year by evaluating OCT features and exploring early toxicity changes in HCQ users. This observational study includes regular image collection and monitoring to improve early diagnosis and management of HCQ maculopathy, with participation lasting about one year.