Clinical Report: AI-Assisted Endoscopy and Secondary Review in Gastric Screening

Overview

This retrospective multicenter before–after study examined whether introducing AI-assisted endoscopy changed secondary image review and referral decisions in a mobile gastric cancer screening program in Gifu Prefecture, Japan. The supplied material describes the study design and cohort of 5,013 examinations but does not provide outcome results, so effects on review or referral decisions cannot be determined from the available information.

Background

Population-based EGD screening is widely conducted in East Asia, including Japan and South Korea, where gastric cancer incidence remains high. In Gifu Prefecture, mobile endoscopy units extend screening access to municipalities, including areas with limited medical resources. The program uses structured secondary review, in which another endoscopist independently examines recorded images and reports before referral is finalized; this supports quality assurance but requires additional personnel and workload. AI-assisted endoscopy has shown lesion-detection potential in controlled and real-time settings, but its effect on secondary review and referral decisions in routine screening remains uncertain.

Data Highlights

Key Findings

• The study included 5,013 consecutive screening EGDs with complete image documentation and clinical data: 2,021 before AI implementation and 2,992 during the AI period.
• CAD EYE was introduced in 2023 to provide real-time visual alerts for regions suspected of gastric neoplastic lesions or esophageal squamous cell carcinoma.
• The AI system was not intended to detect every upper gastrointestinal lesion and did not automatically assign diagnostic classifications or referral decisions.
• Primary endoscopists retained responsibility for initial interpretation and referral categorization, followed by the same structured secondary-review workflow used in the pre-AI period.
• Primary endoscopists were classified as nonexperts if they had under 10 years of experience; experts included board-certified specialists and practitioners with at least 10 years of endoscopic experience.
• The supplied study material does not report comparative outcome data for secondary-review findings, referral decisions, or differences by endoscopist experience. A related report summarized in the supplied context describes a multicenter randomized trial in which AI assistance did not significantly improve gastric-neoplasm detection after centralized pathology review, while reducing blind spots and increasing inspection time.

Clinical Implications

The described workflow retained independent human secondary review after AI-assisted primary interpretation, and the AI system did not make referral decisions. Because comparative results are not included in the supplied study material, no conclusion about whether AI reduced secondary-review workload or altered referral decisions is supported here.

Conclusion

This study evaluated AI implementation within a real-world mobile gastric screening program while maintaining structured human secondary review. The available material establishes the study design and cohort but does not provide the results needed to assess the effects of AI on review or referral outcomes.

Related Resources & Content

1. Study article, title and publication details not provided, Year not provided -- Effects of AI-Assisted Endoscopy on Secondary Image Review and Referral Decisions in Gastric Cancer Screening: A Multicenter Real-World Study
2. The New Gastroenterologist, 2026 -- AI system offers no benefit in upper endoscopy, study finds
3. Korean Practice Guidelines for Gastric Cancer 2024, 2024 -- An Evidence-based, Multidisciplinary Approach (Update of 2022 Guideline)
4. Effectiveness of endoscopic screening for gastric cancer: The Japan Public Health Center-based Prospective Study, publication year not provided -- Effectiveness of endoscopic screening for gastric cancer
5. ASCO Publications — Artificial intelligence-assisted endoscopic diagnosis of upper gastrointestinal neoplasia: A systematic review and meta-analysis of randomized controlled trials and diagnostic test accuracy studies.
6. Surgical Endoscopy — AI Support Reduces Disparities in Endoscopic Reporting of Gastric Lesions: Results from a Prospective Clinical Study
7. The New Gastroenterologist — The Role of Artificial Intelligence in Gastroenterology and Hepatology
8. Performance measures for upper gastrointestinal endoscopy: ESGE Quality Improvement Initiative – Update 2025
9. Effectiveness of endoscopic screening for gastric cancer: The Japan Public Health Center-based Prospective Study
10. Korean Practice Guidelines for Gastric Cancer 2024: An Evidence-based, Multidisciplinary Approach
11. British Society of Gastroenterology guidelines on the diagnosis and management of patients at risk of gastric adenocarcinoma | Gut
12. Efficacy of artificial intelligence–assisted upper gastrointestinal endoscopy for neoplasm detection: a systematic review and meta-analysis of randomized controlled trials - Gastrointestinal Endoscopy
13. Effectiveness of an intelligent endoscopic assistant device for detecting gastric neoplasms and its application conditions: a multi-center, randomized, controlled trial - ScienceDirect
14. World Endoscopy Organization Position Statements for Artificial Intelligence in Endoscopic Diagnosis of Gastric Epithelial Neoplasia - PMC