Artificial Intelligence in Endoscopy for Predicting Helicobacter pylori Infection: A Systematic Review and Meta-Analysis

Purpose

This meta-analysis aimed to assess the diagnostic performance of artificial intelligence (AI) based on endoscopy for detecting Helicobacter pylori (H. pylori) infection.

Methods

A comprehensive literature search was conducted across PubMed, Embase, and Web of Science to identify relevant studies published up to January 10, 2025. The selected studies focused on the diagnostic accuracy of AI in detecting H. pylori. A bivariate random-effects model was employed to calculate pooled sensitivity and specificity, both presented with 95% confidence intervals (CIs). Study heterogeneity was assessed using the I² statistic.

Results

Of 604 studies identified, 16 studies (25,002 images or patients) were included. For the internal validation set, the pooled sensitivity, specificity, and area under the curve (AUC) for detecting H. pylori were 0.91 (95% CI: 0.84–0.95), 0.91 (95% CI: 0.86–0.94), and 0.96 (95% CI: 0.94–0.97), respectively. For the external validation set, the pooled sensitivity, specificity, and AUC were 0.91 (95% CI: 0.86–0.95), 0.94 (95% CI: 0.90–0.97), and 0.98 (95% CI: 0.96–0.99). For junior clinicians, the pooled sensitivity, specificity, and AUC were 0.76 (95% CI: 0.66–0.83), 0.75 (95% CI: 0.70–0.80), and 0.81 (95% CI: 0.77–0.84). For senior clinicians, the pooled sensitivity, specificity, and AUC were 0.81 (95% CI: 0.74–0.86), 0.89 (95% CI: 0.86–0.91), and 0.92 (95% CI: 0.90–0.94).

Conclusions

Endoscopy-based AI demonstrates higher diagnostic performance compared to both junior and senior endoscopists. However, the high heterogeneity among studies limits the strength of these findings, and further research with external validation datasets is necessary to confirm the results.