Artificial Intelligence-assisted Colonoscopy for Detection of Colon Polyps

Not Applicable

• Conditions: Colonoscopy; Artificial Intelligence
• Interventions: Device: Artificial intelligence assisted colonoscopy

• Registration Number: NCT04126265
• Lead Sponsor: Side Liu

Brief Summary

All subjects shall sign informed consent before screening, and subjects shall be included according to inclusion and exclusion criteria.

A total of four endoscopists were included in the study, two in each group of senior endoscopists and two in each group of junior endoscopists.

Patients were randomly enrolled into the senior endoscopy group and the junior endoscopy group, and received artificial intelligence assisted colonoscopy and conventional colonoscopy successively. The two colonoscopy methods were performed back to back by different endoscopy physicians with the same seniority.

All patients were examined and treated according to routine medical procedures. The routine colonoscopy group and the artificial-intelligence-assisted colonoscopy group made detailed records of the patients' withdrawal time, entry time, number of polyps detected, polyp Paris classification, polyp size, polyp shape, polyp location and intestinal preparation during the colonoscopy process

Detailed Description

This is a prospective randomized clinical study.This study was conducted in the Endoscopy Center of the Nanfang Hospital, China. Routine bowel preparation consisted of 4 L of polyethylene glycol, given in split doses. Colonoscopies were performed with high definition colonoscopes and high-definition monitors.

All subjects shall sign informed consent before screening, and subjects shall be included according to inclusion and exclusion criteria.

A total of four endoscopists were included in the study, two in each group of senior endoscopists (\>1000 colonoscopies) and two in each group of junior endoscopists ( \<1000 colonoscopies).

Patients were randomly enrolled into the senior endoscopy group and the junior endoscopy group, and received artificial intelligence assisted colonoscopy and conventional colonoscopy successively. The two colonoscopy methods were performed by different endoscopy physicians back to back with the same seniority.

All patients were examined and treated according to routine medical procedures (outpatient patients and inpatients who did not sign the consent form for polypectomy were not resected for the lesions detected during the examination, while inpatients who signed the consent form for polypectomy were left in the original position after the first colonoscopy and removed at the end of the second examination).

The routine colonoscopy group and the artificial-intelligence-assisted colonoscopy group made detailed records of the patients' withdrawal time, entry time, number of polyps detected, polyp Paris classification, polyp size, polyp shape, polyp location and intestinal preparation during the colonoscopy process.

Study Timeline

• Study Start: 2019-09-01
• Study First Submitted: 2019-10-11
• Study First Submitted QC: 2019-10-14
• Study First Posted: 2019-10-15
• Status Verified: 2019-12
• Last Update Submitted: 2019-12-30
• Last Update Posted: 2020-01-02
• Primary Completion: 2020-04-30
• Study Completion: 2020-08-31

Recruitment & Eligibility

• Status: UNKNOWN
• Sex: All
• Target Recruitment: 560

Inclusion Criteria

Chinese population aged 18-80 years old; Patients voluntarily signed informed consent form; In accordance with the indications of colonoscopy.

Exclusion Criteria

(IBD) history of inflammatory bowel disease; History of colorectal surgery; Previous failed colonoscopy; Polyposis syndrome; Highly suspected colorectal cancer (CRC)

Study & Design

• Study Type: INTERVENTIONAL
• Study Design: PARALLEL

Arm && Interventions

Group	Intervention	Description
Artificial intelligence assisted colonoscopy group	Artificial intelligence assisted colonoscopy	The real-time automatic polyp detection system was used to assist the endoscopist.

Primary Outcome Measures

Name	Time	Method
Detection rate of small polyps (diameter < 6mm)	6 months	In each group, the number of patients with small polyps was detected as a percentage of the total number of patients.

Secondary Outcome Measures

Name	Time	Method
Number of polyps detected	6 months	Number of polyps found in each group.
Polyp size	6 months	Average size of all polyps detected in each group.
Polyp morphology	6 months	Morphological classification of all polyps detected in each group.

Trial Locations

Locations (1)

Nanfang Hospital; 🇨🇳 Guangzhou, Guangdong, China