Work

Artificial intelligence-driven personalized dietary recommendations for gastric cancer high-risk populations: a narrative review

Year: 2026

Type: review

Abstract: This review summarizes the current applications of artificial intelligence (AI) in providing personalized dietary recommendations, and explores its potential applicability to populations at high risk for gastric cancer. Currently, there are no direct intervention trials for gastric cancer patients. However, evidence from metabolic diseases (like diabetes and obesity) shows that AI-driven dietary interventions could be beneficial. This approach may offer translatable benefits for cancer prevention. First, the paper elaborates on the severe incidence of gastric cancer and the limitations of traditional preventive measures, emphasizing the necessity of developing precise and efficient intervention strategies. Subsequently, it systematically outlines methods for identifying high-risk populations and risk stratification (including pathological basis, biomarkers, and genetic risks), as well as the close relationship between dietary patterns (protective and risky) and gastric cancer risk, with a particular focus on the interaction between diet and the gastric microbiome (especially Helicobacter pylori ). The core section analyzes the technical principles of AI-driven personalized nutritional interventions (such as machine learning and deep learning) and their practical effects in improving chronic diseases like blood glucose control and obesity management, while looking forward to the potential of integrating AI with multi-omics data. In addition, the paper extends the discussion to the extended applications of AI in improving screening adherence, assisting endoscopic diagnosis, and clinical decision support systems. Finally, the paper points out current challenges such as technical interpretability, data privacy, population differences, and clinical validation, and proposes prospects for future research directions.

Source: Frontiers in Nutrition

DOI: 10.3389/fnut.2026.1802970