Stanford Medicine: AI Can Be Used to Improve Nutritional Care for Premature Infacts

A study from Stanford Medicine, published in *Nature Medicine*, reveals that artificial intelligence can improve intravenous nutrition administration for premature infants by accurately predicting nutrient needs based on electronic medical records. The AI algorithm, developed using data from nearly 80,000 prescriptions, aims to minimize medical errors and streamline the complex and error-prone process of prescribing total parenteral nutrition (TPN), which is often challenging in neonatal intensive care settings. This innovation is particularly significant as TPN is a major source of medical errors in these units, and the AI tool's ability to assess nutritional requirements for non-communicative infants could enhance overall patient safety and care quality.

Opens on Stanford University