Giordano Antonio, Romero-Mascarell Cristina, González-Suárez Begoña, Guarner-Argente Carlos
Institut de Recerca Sant Pau (IR SANT PAU), Sant Quintí 77-79, 08041, Barcelona, Catalonia, Spain.
Digestive Disease Department, Hospital de la Santa Creu i Sant Pau, Barcelona, Catalonia, Spain.
Dig Dis Sci. 2025 Jun 9. doi: 10.1007/s10620-025-09099-4.
The integration of artificial intelligence (AI) into capsule endoscopy is transforming gastrointestinal diagnostics by enhancing lesion detection and optimizing reading efficiency. This review focuses on the clinical applications of commercially available AI-powered capsule endoscopy systems, particularly for small bowel evaluation. Recent clinical trials and observational studies are analyzed to assess the diagnostic performance, practical benefits, and limitations of these systems. Additionally, key challenges related to standardization, data quality, and clinical validation are discussed. Currently available AI systems significantly reduce reading times and demonstrate high detection capabilities, depending on the algorithm and device used. However, a substantial number of lesions remain undetected, preventing full reliance on these tools. Future advancements must focus on improving detection rates and validating the clinical relevance of missed lesions. Additionally, standardizing AI algorithms across different capsule systems is essential to ensure consistency, reliability, and broader clinical adoption. Establishing homologation frameworks will be key to achieving uniform performance and seamless integration into routine practice.
将人工智能（AI）集成到胶囊内镜检查中，正在通过增强病变检测和优化阅片效率来改变胃肠道诊断。本综述聚焦于市售人工智能驱动的胶囊内镜系统的临床应用，尤其是用于小肠评估。分析了近期的临床试验和观察性研究，以评估这些系统的诊断性能、实际益处和局限性。此外，还讨论了与标准化、数据质量和临床验证相关的关键挑战。目前可用的人工智能系统根据所使用的算法和设备，可显著减少阅片时间并展现出较高的检测能力。然而，仍有大量病变未被检测到，这使得无法完全依赖这些工具。未来的进展必须聚焦于提高检测率以及验证漏诊病变的临床相关性。此外，在不同胶囊系统之间标准化人工智能算法对于确保一致性、可靠性以及更广泛的临床应用至关重要。建立认证框架将是实现统一性能并无缝融入常规实践的关键。
