Artificial intelligence-driven gastrointestinal functional assessment: multimodal imaging, digital biomarkers, and real-time monitoring

Gastrointestinal (GI) functional disorders and chronic inflammatory diseases impose a substantial health burden, yet their assessment remains challenging because symptoms reflect dynamic interactions among motility, visceral sensation, immune–microbiome regulation, and brain–gut signaling. Artificial intelligence (AI) is rapidly reshaping GI functional medicine by enabling scalable, quantitative interpretation of complex data generated from multimodal imaging, physiological sensing, and real-world patient monitoring. This review synthesizes advances across three tightly connected pillars that map onto a physiology-informed “assessment-to-action” loop: (i) AI-assisted multimodal GI imaging for quantitative phenotyping and integrated diagnosis; (ii) AI-enabled discovery and validation of digital biomarkers that capture dynamic GI function in naturalistic settings; and (iii) real-time monitoring platforms that support early warning, longitudinal assessment, and adaptive management. We summarize representative applications in functional GI disorders, inflammatory bowel disease (IBD), and GI oncology, highlighting methodological themes including multimodal fusion, temporal modeling, uncertainty estimation, and explainable AI. We then discuss barriers to translation—standardization and interoperability, external validation under dataset shift, privacy and governance, and workflow integration—and outline practical directions for building clinically trustworthy AI systems for GI functional assessment. Collectively, physiology-centered AI approaches have the potential to transform GI care from episodic testing to longitudinal, mechanism-aware monitoring and personalized intervention.