Artificial Intelligence–Assisted Endoscopy in Diagnosis of Gastrointestinal Tumors: A review of Systematic Reviews and Meta-Analyses

Abstract

AI-assisted endoscopy has emerged as a promising tool for early and accurate detection of gastrointestinal (GI) tumors, which are associated with high morbidity, mortality, and financial burden. This review summarizes systematic reviews and meta-analyses on AI-assisted endoscopy in GI tumor diagnosis. A comprehensive search was conducted using PubMed/MEDLINE, Google Scholar, DOAJ, AJOL, and the Cochrane Library, supplemented by manual searches. Eligible systematic reviews and meta-analyses were selected based on predefined inclusion criteria, and relevant data were extracted to evaluate AI-assisted endoscopy’s diagnostic performance. Out of 569 identified studies, 23 systematic reviews with meta-analyses met the inclusion criteria, with 6 focusing on detection rates and 17 on diagnostic accuracy. AI-assisted endoscopy demonstrated a significantly higher detection rate for GI tumors compared to conventional endoscopy, alongside high diagnostic accuracy across different GI tumor types. However, variability in performance was observed among different AI algorithms and studies. While AI-assisted endoscopy enhances diagnostic precision, rigorous validation of AI models is necessary to ensure clinical reliability. Ethical considerations and further research are crucial for optimizing AI’s role in healthcare.