Application of Artificial Intelligence on the Diagnosis of Helicobacter Pylori Infection and Premalignant Gastric Lesion

Application of Artificial Intelligence on the Diagnosis of Helicobacter Pylori Infection and Premalignant Gastric Lesion: A Randomized Clinical Trial

The aim of this study is to evaluate the impact of artificial intelligence (AI) assistance during routine upper endoscopy on gastric cancer-specific mortality. We hypothesize that AI-assisted endoscopic interpretation can further reduce gastric cancer-related mortality through two mechanisms: (1) improved detection of H. pylori infection, facilitating timely eradication therapy and subsequent prevention of gastric carcinogenesis; and (2) earlier identification of premalignant gastric conditions, enabling appropriate surveillance endoscopy and earlier detection of gastric cancer. The primary endpoint is gastric cancer-specific mortality.

Study Overview

• Status: Recruiting
• Conditions: Helicobacter Pylori Infection; Premalignant Lesion
• Intervention / Treatment: Other: Routine endoscopy with artificial intelligence-assisted interpretation

Detailed Description

This is a randomized clinical trial designed to evaluate the effectiveness of AI-assisted endoscopy compared with routine endoscopy in reducing gastric cancer-specific mortality. Participants will be allocated in a 1:1 ratio using a computer-generated randomization sequence after eligibility assessment and informed consent. One group will receive artificial intelligence-assisted interpretation for physician reference, while the control group will undergo routine endoscopy without AI assistance. In routine clinical practice, patients diagnosed with H. pylori infection receive antibiotic eradication therapy to reduce the risk of gastric cancer incidence and mortality, while those with premalignant gastric conditions are generally advised to undergo surveillance upper endoscopy every two years. We hypothesize that AI-assisted interpretation may further reduce gastric cancer-specific mortality through two mechanisms: (1) improved detection of H. pylori infection, enabling timely eradication therapy; and (2) earlier identification of gastric cancer via AI-supported detection of premalignant gastric conditions and appropriate recommendations for surveillance endoscopy. The primary endpoint is gastric cancer-specific mortality.

• Study Type: Observational
• Enrollment (Estimated): 6000

Contacts and Locations

• Study Contact: Name: Tsung-Hsien Chiang, MD,PhD; Phone Number: 265427 886-2-23123456; Email: thchiang@ntu.edu.tw
• Study Contact Backup: Name: Yi-Chia Lee, MD, PhD; Phone Number: 265689 886-2-23123456; Email: yichialee@ntu.edu.tw

Study Locations

• Taiwan
  • Taipei, Taiwan, 10015
    • Recruiting
    • Yi-Chia Lee
    • Contact: Yi-Chia Lee; Email: yichialee@ntu.edu.tw

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 20 years to 80 years (Adult, Older Adult)
• Accepts Healthy Volunteers: No

• Sampling Method: Non-Probability Sample

Study Population

This study will invite patients who need to undergo upper gastrointestinal endoscopy.

Description

Inclusion Criteria:

1. Age 20-80
2. Scheduled endoscopy

Exclusion Criteria:

1. History of gastric surgery

Study Plan

How is the study designed?

Number of groups / cohorts

2

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
Routine endoscopy alone; Routine endoscopy without artificial intelligence assistance
Routine endoscopy with artificial intelligence-assisted interpretation; Routine endoscopy assisted by artificial intelligence to enhance the detection of H. pylori infection and premalignant gastric conditions.	Other: Routine endoscopy with artificial intelligence-assisted interpretation; (1) Improved detection of H. pylori infection, leading to timely eradication therapy. (2) Earlier identification of premalignant gastric conditions, facilitating appropriate surveillance endoscopy.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Gastric cancer-specific mortality	The primary endpoint is gastric cancer-specific mortality.	Up to 5 years

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Specificity to exclude premalignant gastric conditions	Outcomes include the atrophic gastritis, intestinal metaplasia, and Helicobacter pylori infection	Up to 5 years

Other Outcome Measures

Outcome Measure	Measure Description	Time Frame
Areas under the ROC curves	Outcomes include the atrophic gastritis, intestinal metaplasia, and Helicobacter pylori infection	Up to 5 years
Gastric cancer incidence	New incident gastric cancer	Up to 10 years

Collaborators and Investigators

• Sponsor: National Taiwan University Hospital
• Investigators: Principal Investigator: Tsung-Hsien Chiang, MD, PhD, National Taiwan University Hospital

Study record dates

Study Major Dates

• Study Start (Actual): December 24, 2021
• Primary Completion (Estimated): December 31, 2028
• Study Completion (Estimated): December 31, 2028

Study Registration Dates

• First Submitted: February 28, 2023
• First Submitted That Met QC Criteria: February 28, 2023
• First Posted (Actual): March 10, 2023

Study Record Updates

• Last Update Posted (Actual): June 24, 2026
• Last Update Submitted That Met QC Criteria: June 21, 2026
• Last Verified: June 1, 2026

More Information

Terms related to this study

• Keywords: Helicobacter, Artificial Intelligence, Telemedicine
• Additional Relevant MeSH Terms: Pathologic Processes; Neoplasms; Disease Attributes; Bacterial Infections; Bacterial Infections and Mycoses; Gram-Negative Bacterial Infections; Infections; Communicable Diseases; Helicobacter Infections; Precancerous Conditions
• Other Study ID Numbers: 202111108RINC

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: UNDECIDED
• IPD Plan Description: The study was based on the images and histological data.

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No
• product manufactured in and exported from the U.S.: No