Project Scope:  This project aims to develop a multimodal AI-based system for detecting and predicting refractive errors in a worldwide population (i.e., starting with Ghana and France) by integrating ocular appearance images with clinical and lifestyle data. Unlike traditional AI models for refractive error detection that rely solely on retinal or corneal imaging, our approach will incorporate environmental and behavioral risk factors, such as close-up work, educational level, outdoor activity, parental ocular history, and personality traits (e.g., introversion), to increase the accuracy of the system. Additionally, other quantitative biological signs, such as temperature and light refraction, will be incorporated into the sensor-detected data. The project aims to improve early detection and screening for refractive errors, particularly in regions with limited access to eye care services.

Publications and open-access (data and codes) links: