Human milk composition exhibits significant temporal and inter-individual variability driven by maternal, infant, and environmental factors. Traditional analytical methods are limited in capturing the complex non-linear relationships governing milk composition and lactation outcomes. To systematically evaluate the applications of artificial intelligence (AI) and machine learning (ML) in human milk analysis, donor milk optimization, and lactation support. Comprehensive search of PubMed/MEDLINE, Embase, Cochrane Library, Scopus, and Web of Science (2015–2025) for studies applying AI/ML to human milk research and breastfeeding support. AI/ML models achieved 87.9% accuracy (AUC 0.917) for predicting low milk supply and R²>0.85 for macronutrient prediction. AI-optimized donor milk pooling improved nutritional target achievement by 31% and reduced processing time by 60%. Convolutional neural networks demonstrated AUC 0.93–0.96 for detecting breastfeeding complications. AI chatbots significantly improved breastfeeding self-efficacy scores (p = 0.027). Multi-omics integration revealed complex temporal patterns in milk composition. AI/ML technologies demonstrate transformative potential for personalized maternal-infant care through enhanced prediction accuracy, operational efficiency, and accessible lactation support. Clinical translation requires addressing methodological standardization, diverse population validation, and model interpretability.
Golshan-Tafti et al. (Tue,) studied this question.