In Kenya, 350,000 children suffer from acute malnutrition, but current forecasting often misses rapid surges. USC and Microsoft researchers developed an AI model that integrates data from 17,000 health clinics with satellite imagery of crop health. The tool predicts malnutrition up to six months early with 86% accuracy. This provides governments with a critical head start to deliver life-saving food and care, preventing unnecessary deaths and better protecting at-risk communities.
