A rural-to-center artificial intelligence model for diagnosing Helicobacter pylori infection and premalignant gastric conditions using endoscopy images captured in routine practice

Diagnosing <i>Helicobacter pylori</i> infection and premalignant gastric conditions typically requires ¹³C urea breath testing or histological assessment, which are often unavailable in remote areas. A rural-to-center artificial intelligence (AI) model was developed and implemented to automatically evaluate upper endoscopy images from routine clinical practice.Endoscopic images were collected from a rural hospital on Matsu Islands and a tertiary center across Taiwan Strait. During model development (2020-2022), AI algorithms were trained, validated, and tested to exclude low-quality and non-gastric images, segment gastric regions, and enhance mucosal features for detecting <i>H. pylori</i> infection and premalignant conditions. During model implementation (2023-2024), endoscopic images from a rural hospital were transmitted to the medical center for AI analyses, with results promptly returned.In the development phase, diagnostic accuracies were 92.8% (95%CI 88.9%-96.6%) for <i>H. pylori</i>, 88.6% (95%CI 87.2%-90.0%) for atrophic gastritis, and 88.0% (95%CI 86.5%-89.5%) for intestinal metaplasia. In the implementation phase, 3518 rural residents underwent ¹³C urea breath testing or pepsinogen testing; 421 with positive results underwent endoscopy. No significant differences were observed between AI-predicted and clinically observed prevalence: <i>H. pylori</i> (13.9% vs. 12.9%; <i>P</i> = 0.55), atrophic gastritis (15.7% vs. 11.9%; <i>P</i> = 0.34), and intestinal metaplasia (27.6% vs. 22.4%; <i>P</i> = 0.32). Implementation-phase diagnostic accuracies were 91.3% (95%CI 88.0%-94.6%), 79.9% (95%CI 72.1%-86.3%), and 63.4% (95%CI 54.7%-71.6%), respectively.AI enabled physicians in resource-limited settings to rapidly assess gastric health using routinely captured endoscopic images, bridging gaps in access and expertise.