Abstract
Purpose of review
High-resolution anorectal manometry (HRAM) has improved anorectal physiological assessment but remains limited by conventional pressure-based metrics that incompletely reflect the complex spatiotemporal physiology of continence and defecation. This review summarizes recent advances in artificial intelligence (AI) and machine learning and their potential to enhance HRAM interpretation and clinical decision-making.Recent findings
Recent AI-based approaches have shown promising performance in automated HRAM interpretation, prediction of dyssynergic defecation and balloon expulsion abnormalities, implementation of the London Classification, and analysis of three-dimensional high-definition anorectal manometry datasets. In parallel, integrated physiological metrics such as integrated pressurized volume (IPV) and time-series IPV (TS-IPV) have improved characterization of anorectal function by incorporating pressure amplitude, spatial distribution, and temporal dynamics. Emerging deep-learning models have further demonstrated the potential to identify novel physiological patterns beyond traditional classification systems. AI has the potential to reduce observer variability, improve diagnostic standardization, and facilitate outcome prediction in anorectal physiology testing. Although challenges related to data standardization, external validation, and explainable AI remain, the integration of advanced physiological metrics and AI-driven analytics may transform HRAM from a descriptive diagnostic test into a platform for dynamic physiological phenotyping and precision medicine.