When a parent’s own milk is not available, pasteurized donor human milk is the preferred supplement instead of formula as it can lower the risk of a potentially life-threatening bowel condition in preterm infants. The Rogers Hixon Ontario Human Milk Bank is one of three accredited Milk Banks in Canada. We screen donors (questionnaire, blood test), collect milk (by courier), pool milk from multiple donors to create batches of donor milk. To improve our operations and to ensure the quality of our donor milk product, we present a novel AI-driven approach that combines machine learning and optimization to predict macronutrient content and support evidence-based pooling decisions. By improving efficiency and increasing consistency in meeting clinical targets, this work demonstrates how AI can be a practical tool to enhance care delivery in milk banking and other healthcare settings.
Moderator: Kyla Reimer, CHIM