AI-Enabled Precision Nutrition in the ICU: A Narrative Review and Implementation Roadmap

Area of Science:

• Critical care medicine
• Biomedical informatics
• Nutritional science

Background:

• Artificial intelligence (AI) is transforming intensive care units (ICUs) with personalized care and real-time analytics.
• Nutritional therapy, crucial for ICU outcomes, often faces delays and lacks individualization.

Purpose of the Study:

• To review current and emerging AI applications in ICU nutrition.
• To highlight the clinical potential, implementation barriers, and ethical considerations of AI in critical care nutrition.

Main Methods:

• A narrative review of English-language literature from January 2018 to November 2025.
• Searches in PubMed/MEDLINE, Scopus, Web of Science, Google Scholar, and citation tracking.
• Focused on machine learning (ML), deep learning (DL), natural language processing (NLP), and reinforcement learning (RL) for nutrition optimization.

Main Results:

• AI models can accurately estimate nutritional needs and optimize enteral (EN) and parenteral nutrition (PN).
• AI predicts gastrointestinal intolerance and metabolic complications, enabling real-time therapy adaptation.
• Reinforcement learning (RL) and multi-omics integration facilitate precision nutrition through longitudinal data analysis.

Conclusions:

• High-quality data, oversight, and clinician education are essential for AI implementation in ICU nutrition.
• AI can enhance human expertise, leading to safer and more targeted nutrition delivery.
• Prioritizing transparency, equity, and workflow integration is key for successful AI adoption.