Evaluation of the Efficacy and Safety of an AI-Based Endoscopic Image Analysis Software During Real-Time Upper Gastrointestinal Endoscopy

Evaluate the Efficacy and Safety of an Artificial Intelligence-Based Endoscopic Image Analysis Software During Real-Time Upper Gastrointestinal Endoscopic Examinations: A Prospective, Single-Center, Post-Market Exploratory Clinical Study

Status

Active, not recruiting

Phases

Unknown

Study type

Interventional

Source

CRIS

Registry ID

KCT0012411

Enrollment

100

Registered

2026-08-06

Start date

2026-02-09

Completion date

Unknown

Last updated

2026-08-10

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

None listed

Interventions

Medical Device : The investigational device (MD-GA-300) is an endoscopic image analysis software that displays abnormal areas (such as elevated or depressed lesions) on endoscopic images in real time.

Eligibility

Sex/Gender

All

Inclusion criteria

Inclusion criteria: - Subjects who voluntarily agreed to participate and signed the written informed consent form - Adults aged 19 years or older scheduled to undergo endoscopy

Exclusion criteria

Exclusion criteria: - Patients undergoing emergency endoscopy - Cases in which gastric visualization is not possible due to foreign bodies or patient hypersensitivity - Patients with anatomical alterations of the stomach

Design outcomes

Primary

Measure	Time frame
Lesions per EGD (overall):The number of all lesions observed in the stomach during upper GI endoscopy. (erosion, depression, polyp, elevated lesion, gastric ulcer of any stage, mass, cancer)	—

Secondary

Measure	Time frame
Lesions per EGD (significant):The number of lesions among all observed lesions that were judged to require biopsy or that actually underwent biopsy.;Lesions per EGD (pathologic): The number of lesions confirmed as neoplastic findings (atypia, dysplasia of any grade, cancer) on biopsy;Validation of endoscopic videos from the AI-assisted novice group;Examiner satisfaction;Examiner workload (NASA-TLX);Safety Evaluation: Presence or absence of adverse events and complications	—

Measure

Time frame

Lesions per EGD (significant):The number of lesions among all observed lesions that were judged to require biopsy or that actually underwent biopsy.;Lesions per EGD (pathologic): The number of lesions confirmed as neoplastic findings (atypia, dysplasia of any grade, cancer) on biopsy;Validation of endoscopic videos from the AI-assisted novice group;Examiner satisfaction;Examiner workload (NASA-TLX);Safety Evaluation: Presence or absence of adverse events and complications

—

Countries

Korea, Republic of

Contacts

Public ContactSang Kil Lee

Yonsei University Health System, Severance Hospital

qndrj@naver.com+82-2-2228-4000

Outcome results