The Improvement Effect of Real-time Artificial Intelligence Assisted Identification of Bleeding Points on Hemostasis Efficiency in Endoscopic Submucosal Dissection

Shandong University

Status: Not yet enrolling

Conditions: Endoscopic Submucosal Dissection (ESD); Endoscopic Submucosal Dissection

Treatments: Device: AI real-time assistance in endoscopic submucosal dissection (ESD) for bleeding spot identification and marking

Study type: Interventional

Funder types: Other

Identifiers

NCT07495137

2026SDU-QILU-1

Details and patient eligibility

About

The goal of this clinical trial is to learn if an artificial intelligence (AI) system that identifies bleeding points in real time can help stop bleeding faster during endoscopic submucosal dissection (ESD) - a minimally invasive surgery for early digestive tract cancer or precancerous lesions. It will also learn about the AI system's effect on surgery-related problems (like perforation or delayed bleeding) and total surgery time.

The main questions it aims to answer are:

1. Does the AI system shorten the time it takes to stop each bleed during ESD?
2. How does the AI system affect the rate of surgery-related problems and total surgery time?

Researchers will compare two groups to see if the AI system improves hemostasis efficiency:

1. AI group: During ESD, the AI system will real-time spot and mark bleeding points. Doctors will use these marks to stop bleeding.
2. Control group: Doctors will use the same equipment but without the AI system - they will find and stop bleeding using their own experience.

Participants will:

1. Have ESD surgery for esophageal, stomach, or colorectal lesions that need this treatment;
2. Be randomly assigned to either the AI group or the control group;
3. Attend follow-up checks in 14 days after surgery to check for complications;
4. Have their surgery videos reviewed by experts to record hemostasis time and total surgery time.

Enrollment: 160 estimated patients

Sex: All

Ages: 18 to 80 years old

Volunteers: No Healthy Volunteers

Inclusion and exclusion criteria

Inclusion Criteria

1. Aged 18-80 years;
2. Lesions meet the indications for ESD treatment of the esophagus, stomach, or colorectum according to relevant guidelines;
3. Anticoagulant drugs have been suspended according to relevant guidelines;
4. Patients with American Society of Anesthesiologists (ASA) classification Grade I or II;
5. Patients who voluntarily sign the informed consent form.

Exclusion Criteria

1. Patients with severe cardiopulmonary diseases, coagulation dysfunction or other severe comorbidities that may increase surgical risks;
2. Patients undergoing dialysis treatment;
3. Pregnant or lactating women;
4. Deemed unsuitable for participation in this study by the principal investigator or other researchers.

Trial design

Primary purpose: Treatment

Allocation: Randomized

Interventional model: Parallel Assignment

Masking: Triple Blind

---

160 participants in 2 patient groups

AI-assisted Group

Experimental group

Description:

During the endoscopic submucosal dissection (ESD) procedure, an artificial intelligence (AI) real-time bleeding point recognition system is utilized. The system dynamically analyzes endoscopic images to identify and mark bleeding sites in real time. Endoscopists perform hemostatic operations promptly based on these AI-generated markers.

Treatment:

Device: AI real-time assistance in endoscopic submucosal dissection (ESD) for bleeding spot identification and marking

Conventional treatment group

No Intervention group

Description:

Patients undergo ESD using the same hardware platform, but the AI system is deactivated. Hemostatic decisions and operations are solely dependent on the endoscopists' clinical experience-they independently judge the location of bleeding points and perform hemostasis without AI assistance.