Automated Quantitative Ulcer Analysis: Diagnosing Fungal Organisms for Corneal Ulcers

This project develops AQUA (Automated Quantitative Ulcer Analysis), a free, open-source deep-learning algorithm to help clinicians in low-resource settings identify fungal organism types in microbial keratitis — a leading cause of blindness that disproportionately affects agricultural communities in warm-climate LMICs. Built on 1,700+ prospectively collected cases from Aravind Eye Care Systems in India, the algorithm analyzes slit-lamp photography and clinical risk factors to classify fungal subtypes at the point of care, filling a critical gap where current gold-standard lab methods are too slow, expensive, or unavailable. The project delivers both the algorithm and a cloud-based clinical decision tool accessible to under-resourced eye clinics worldwide.