Artificial intelligence to support the definition of nutrition in premature newborns. This is the underlying premise of a study published in the Journal of Perinatology, a "Nature" portfolio journal, as the result of joint research by researchers from the Fondazione IRCCS San Gerardo dei Tintori (FSGT) and the Department of Electronics, Information, and Bioengineering (DEIB) at the Polytechnic University of Milan. The study focuses on one of the most difficult stages in the care of very preterm infants: the shift from intravenous (parenteral) to oral (enteral) feeding. This "nutritional transition" phase is critical for growth and development, but it is currently managed without standardized procedures backed by strong scientific data. Excessive, insufficient, or uneven nutrient intake can lead to problems and growth retardation. "In very premature infants, growth isn't just a numerical indicator: a slowdown in extrauterine growth can have long-lasting consequences, potentially impacting neurocognitive development", the experts explain. "This is why studying nutritional transition involves aiming not only for increased growth but also for improved overall developmental quality. To do so, we must first comprehend what happens during life's most sensitive days, when every decision counts. This is where data has its greatest impact". Artificial intelligence has enabled the integration of vast volumes of heterogeneous clinical data, transforming them into useful research tools and, increasingly, clinical decision assistance.