UMIN-ICDS Clinical Trial

Public title

Study on the Educational Effectiveness of AI-Based Feedback in Upper Gastrointestinal Endoscopy Training

Acronym

AI-Endoscopy Education Study

Scientific Title

Educational Effectiveness of AI-Based Feedback in Early Upper Gastrointestinal Endoscopy Training for Non-Expert Endoscopists: A Non-Inferiority Randomized Controlled Trial Based on Completeness of Mucosal Visualization and Image Quality

Scientific Title:Acronym

AI-UGI Education Non-Inferiority Trial

Region

Japan

Condition

Adults undergoing diagnostic upper gastrointestinal endoscopy

Classification by specialty

Gastroenterology

Classification by malignancy

Others

Genomic information

NO

Narrative objectives1

The objective of this study is to evaluate whether upper gastrointestinal endoscopy training using feedback based on AI-generated analysis reports is non-inferior to conventional supervisor-based feedback in terms of procedural completeness among non-expert endoscopists in the early phase of training.

Basic objectives2

Efficacy

Basic objectives -Others

Trial characteristics_1

Confirmatory

Trial characteristics_2

Pragmatic

Developmental phase

Not applicable

Primary outcomes

The case-level completeness of mucosal visualization score (%) calculated by the same AI algorithm (gastro BASE screening X) after each eligible upper gastrointestinal endoscopy performed during the registration period. The between-period difference in the completeness score between the AI-report period and the supervisor-feedback period will be evaluated.

Key secondary outcomes

For each eligible upper gastrointestinal endoscopy performed during the registration period, the following outcomes will be assessed after the procedure or after AI analysis: procedure time, total number of captured images, number of gastric images, number of gastric image-enhanced endoscopy images, number of gastric chromoendoscopy images, number of biopsies, number of lesions diagnosed as Group 2 or higher, number of images with inadequate image quality, operator-level improvement rate in the completeness score, and inter-operator variability.

Study type

Interventional

Basic design

Cross-over

Randomization

Randomized

Randomization unit

Cluster

Blinding

Open -no one is blinded

Control

Active

Stratification

NO

Dynamic allocation

NO

Institution consideration

Institution is not considered as adjustment factor.

Blocking

YES

Concealment

Central registration

No. of arms

2

Purpose of intervention

Educational,Counseling,Training

Type of intervention

Other

Interventions/Control_1

AI-report period: During weekly or daily time blocks assigned to each participating physician within the registration period, upper gastrointestinal endoscopy will be performed as part of routine clinical practice. After each procedure, coded endoscopic images will be analyzed in the cloud using the same AI algorithm, gastro BASE screening X, and an AI-generated report including the completeness score and image-quality indicators will be provided to the operator. The operator will conduct self-reflection based on brief feedback after each case and summary reviews approximately every five cases. Approximately 32 cases are planned for the AI-report period for each physician.

Interventions/Control_2

Supervisor-feedback period: During weekly or daily time blocks assigned to each participating physician within the registration period, upper gastrointestinal endoscopy will be performed as part of routine clinical practice. After each procedure, AI analysis will be performed in the same manner as in the AI-report period; however, the AI-generated results will not be disclosed to either the operator or the supervisor during the feedback period. The supervisor will provide brief feedback after each case and summary reviews approximately every five cases based on his or her own clinical judgment. Approximately 32 cases are planned for the supervisor-feedback period for each physician.

Interventions/Control_3

Interventions/Control_4

Interventions/Control_5

Interventions/Control_6

Interventions/Control_7

Interventions/Control_8

Interventions/Control_9

Interventions/Control_10

Age-lower limit

20

years-old

<=

Age-upper limit

Not applicable

Gender

Male and Female

Key inclusion criteria

Patient inclusion criteria: Patients aged 20 years or older, regardless of sex, who undergo diagnostic upper gastrointestinal endoscopy. Eligible indications include health check-up, dyspepsia, follow-up after Helicobacter pylori eradication, follow-up of ulcers or erosions, follow-up of chronic gastritis, and evaluation of functional gastrointestinal disorders. Patients should be classified as ASA physical status I or II and be able to undergo endoscopy with standard preparation and sedation. Patients who do not opt out after public disclosure of the study information will be included.
Physician inclusion criteria: Physicians who have not yet obtained board certification from the Japan Gastroenterological Endoscopy Society, have performed approximately 50 to 300 upper gastrointestinal endoscopies independently, and are able to perform the procedure safely. Physicians must be able to perform approximately four cases per week for about 16 weeks and comply with the predefined allocation to the AI-report and supervisor-feedback periods. Written informed consent must be obtained from participating physicians.

Key exclusion criteria

Patient exclusion criteria: Patients undergoing endoscopy for therapeutic or detailed diagnostic purposes, including evaluation of suspected cancer, active bleeding, varices, marked stenosis, or planned EMR or ESD. Patients with anatomical conditions that substantially deviate from the standard observation map, such as post-gastrectomy status or Roux-en-Y reconstruction, will be excluded. Patients classified as ASA physical status III or IV, emergency cases, those with severe comorbidities, contraindications to sedation, or pregnancy will be excluded. Patients who opt out after public disclosure of the study information or withdraw from study participation will also be excluded. Patients judged by the principal investigator or co-investigators to be unsuitable for endoscopy by a beginner will be excluded.
Physician exclusion criteria: Physicians who have already obtained board certification from the Japan Gastroenterological Endoscopy Society or are too highly experienced to allow evaluation of the educational effect will be excluded. Physicians who cannot comply with the allocation during the study period or who do not provide consent will also be excluded.

Target sample size

512

Name of lead principal investigator

1st name	Tadateru
Middle name
Last name	Maehata

Organization

St. Marianna University School of Medicine

Division name

Department of Gastroenterology

Zip code

216-8511

Address

2-16-1 Sugao, Miyamae-ku, Kawasaki, Kanagawa 216-8511, Japan

TEL

044-977-8111

Email

t2maehata@marianna-u.ac.jp

Name of contact person

1st name	Tadateru
Middle name
Last name	Maehata

Organization

St. Marianna University School of Medicine

Division name

Department of Gastroenterology

Zip code

216-8511

Address

2-16-1 Sugao, Miyamae-ku, Kawasaki, Kanagawa 216-8511, Japan

TEL

044-977-8111

Homepage URL

Email

t2maehata@marianna-u.ac.jp

Institute

St. Marianna University School of Medicine

Institute

Department

Personal name

Tadateru Maehata

Organization

St. Marianna University School of Medicine

Organization

Division

Category of Funding Organization

Self funding

Nationality of Funding Organization

Japan

Co-sponsor

Name of secondary funder(s)

Organization

Ethics Committee (Clinical Trials Subcommittee), St. Marianna University School of Medicine

Address

2-16-1 Sugao, Miyamae-ku, Kawasaki, Kanagawa 216-8511, Japan

Tel

044-977-8111

Email

k-sienbu.mail@marianna-u.ac.jp

Secondary IDs

NO

Study ID_1

Org. issuing International ID_1

Study ID_2

Org. issuing International ID_2

IND to MHLW

Institutions

聖マリアンナ医科大学病院（神奈川県)

Date of disclosure of the study information

2026

Year

06

Month

08

Day

URL releasing protocol

Publication of results

Unpublished

URL related to results and publications

Number of participants that the trial has enrolled

Results

Results date posted

Results Delayed

Results Delay Reason

Date of the first journal publication of results

Baseline Characteristics

Participant flow

Adverse events

Outcome measures

Plan to share IPD

IPD sharing Plan description

Recruitment status

Preinitiation

Date of protocol fixation

2026

Year

04

Month

30

Day

Date of IRB

Anticipated trial start date

2026

Year

06

Month

08

Day

Last follow-up date

2026

Year

12

Month

31

Day

Date of closure to data entry

Date trial data considered complete

Date analysis concluded

Other related information

Registered date

2026

Year

05

Month

27

Day

Last modified on

2026

Year

05

Month

27

Day

Link to view the page

Value
https://center6.umin.ac.jp/cgi-open-bin/icdr_e/ctr_view.cgi?recptno=R000070578