<![CDATA[Introduction: Acute gastrointestinal bleeding (GIB) is a life-threatening condition requiring prompt diagnosis and intervention. Despite advancements in endoscopy, challenges like diagnostic variability and resource allocation persist. Artificial intelligence (AI)-assisted tools have shown promise in improving detection rates and workflow efficiency in gastroenterology, but their impact on clinical decision-making and patient outcomes in acute GIB remains underexplored. Methods: This systematic review adhered to PRISMA 2020 guidelines. Eligible studies included those evaluating AI-based diagnostic tools in acute GIB cases, with outcomes such as diagnostic accuracy, clinical decision-making changes, or patient outcomes. Databases like PubMed, Sagepub, and Google Scholar were searched using Boolean MeSH keywords. Data extraction focused on study design, AI tool characteristics, diagnostic performance, and clinical impact. Results: Among 40 included studies, AI demonstrated high diagnostic accuracy, with sensitivities and specificities exceeding 90% in lesion detection. For instance, convolutional neural networks achieved 95.4% accuracy in identifying ulcers and hemorrhages. However, only one study reported a modest improvement in predicting endoscopic intervention needs (AUC: 0.68). While AI reduced reading time by 30% in some studies, its impact on patient outcomes (e.g., mortality, rebleeding) was rarely addressed. Most evidence came from retrospective studies or meta-analyses, with limited prospective or randomized controlled trials. Discussion: AI enhances diagnostic accuracy and workflow efficiency but lacks robust evidence linking it to improved patient outcomes in acute GIB. Key limitations include methodological heterogeneity, scarce safety data, and a focus on non-acute settings. Prospective studies are needed to evaluate AI's real-world clinical impact. Conclusion: AI shows potential as an adjunct tool in acute GIB management but requires further validation to confirm its clinical utility. Future research should prioritize patient-centered outcomes and standardized reporting.]]>
Copyrights © 2026
