臨床研究等提出・公開システム

Japanese

Date of registration	June. 03, 2026
Last modified on	June. 15, 2026
Trial ID	jRCT1032260188

Scientific Title	Diagnostic Performance and Cost-Effectiveness Analysis of AI-Based Anatomical Landmark Recognition in Upper Endoscopy
Public Title	Diagnostic Performance and Cost-Effectiveness Analysis of AI-Based Anatomical Landmark Recognition in Upper Endoscopy

Contact for Scientific Queries	Name	Sonoda Takayoshi
	Affiliation	Cancer Institute Hospital of JFCR
	Address	3-8-31, Ariake, Koto-ku, Tokyo, 135-8550
	Telephone	+81-3-3520-0111
	E-mail	hirotaka.nakashima@jfcr.or.jp
Contact for Public Queries	Name	Sonoda Takayoshi
	Affiliation	Cancer Institute Hospital of JFCR
	Address	3-8-31, Ariake, Koto-ku, Tokyo, 135-8550
	Telephone	+81-3-3520-0111
	E-mail	takayoshi.sonoda@jfcr.or.jp

Recruitment status	Pending

Anticipated date of first enrollment		June. 03, 2026
Target sample size		200
Study Type		Interventional
Study Design	allocation	randomized controlled trial
	masking	single blind
	control	uncontrolled control
	assignment	single assignment
	purpose	diagnostic purpose
Key inclusion & exclusion criteria	Inclusion Criteria	Subjects aged 20 years or older. Individuals undergoing upper gastrointestinal endoscopy (EGD) screening at the Cancer Screening Center, Cancer Institute Hospital of JFCR. Subjects who have provided written informed consent voluntarily after receiving a full explanation regarding participation in this study.
	Exclusion Criteria	Subjects with a history of gastric resection. Individuals with distinct clinical gastrointestinal symptoms. Other individuals judged to be inappropriate for participation in this study by the principal investigator or sub-investigators.
	Age Minimum	20age old over
	Age Maximum	No limit
	Gender	Both
Health Condition(s) or Problem(s) Studied		Gastric Neoplasms
Intervention(s)		Subjects undergoing screening EGD will be randomly assigned to two groups to evaluate the diagnostic performance of "Landmark Photo Checker (LMPC)", an endoscopic AI software function implemented in CAD EYE (FUJIFILM Co.). Sensitivity Validation Group: Endoscopists capture appropriate images of seven anatomical landmarks to prospectively evaluate LMPC's true positive rate. Specificity Validation Group: Endoscopists intentionally capture inappropriate images (obscured or partial views) to evaluate LMPC's true negative rate. The additional imaging procedure takes only a few seconds and does not affect the quality of standard clinical practice.
Health Condition(s) Keyword		Neoplasm, Stomach
Intervention(s) Keyword		Artificial Intelligence, Endoscopic AI, Esophagogastroduodenoscopy, Quality Control, Landmark Photo Checker
Health Condition(s) Code		D013274
Intervention(s) Code		D005773
Primary Outcome(s)		Sensitivity of the endoscopic AI software function, Landmark Photo Checker (LMPC), in automatically judging appropriate imaging for seven anatomical landmarks during esophagogastroduodenoscopy (EGD)
Secondary Outcome(s)		pecificity and accuracy of the endoscopic AI software function, Landmark Photo Checker (LMPC), in automatically judging appropriate imaging for seven anatomical landmarks during esophagogastroduodenoscopy (EGD). Cost-effectiveness (5-year average costs and net monetary benefit: NMB) of implementing LMPC in gastric cancer screening using a microsimulation model based on the clinicopathological characteristics of detected and false-negative gastric cancers

Secondary Sponsor

Secondary Sponsor

Name of Certified Review Board	The Japanese Foundation for Cancer Research, Medical Research Ethics Review Committee
Address	3-8-31 Ariake, Koto-ku, Tokyo, Tokyo
Telephone	+81-3-3520-0111
E-Mail	med.shinsa@jfcr.or.jp
Approval Status	Approval
Date of approval	May. 29, 2026

Plan to share IPD	No

Countries of Recruitment（Except Japan）	none

History of Changes

No	Publication date
2	June. 15, 2026	(this page)	Changes
1	June. 03, 2026	Detail

List of change history