Prospective Cohort Study of the Development of AI Diagnosing Subepithelial Lesions of the Gastrointestinal Tract

Prospective Cohort Study of Collection and Validation of Clinical Data for the Development of AI Diagnosing Subepithelial Lesions of the Gastrointestinal Tract - Development of SEL-AI

Status

Recruiting

Phases

Unknown

Study type

Observational

Source

JPRN

Registry ID

JPRN-UMIN000054285

Enrollment

720

Registered

2024-04-30

Start date

2024-04-30

Completion date

Unknown

Last updated

2026-06-29

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

Conditions

Cases undergoing EUS examination aimed at diagnosis, tissue sampling, and follow-up for suspected submucosal tumors of the gastrointestinal tract

Interventions

None listed

Eligibility

Sex/Gender

All

Inclusion criteria

Inclusion criteria: Cases undergoing upper gastrointestinal EUS examination aimed at precise inspection, tissue sampling, and follow-up for lesions suspected of SEL, with the age of consent being 18 years or older at the time of consent

Exclusion criteria

Exclusion criteria: Patients deemed unsuitable as research subjects by the principal investigator; patients lacking the capacity to consent

Design outcomes

Primary

Measure	Time frame
Accuracy in SEL's three-category differentiation (GIST, leiomyoma, and others) in EUS by an AI diagnostic system	—

Secondary

Measure	Time frame
Diagnostic performances (sensitivity, specificity, positive predictive value, negative predictive value, accuracy) of an AI diagnostic system and expert physicians in the two-category differentiation (GIST, others) of SEL; diagnostic performances in multi-category classification (accuracy, sensitivity per category); diagnostic capabilities of non-expert physicians; diagnostic capabilities of various biopsy procedures; and EUS findings contributing to diagnosis	—

Measure

Time frame

Diagnostic performances (sensitivity, specificity, positive predictive value, negative predictive value, accuracy) of an AI diagnostic system and expert physicians in the two-category differentiation (GIST, others) of SEL; diagnostic performances in multi-category classification (accuracy, sensitivity per category); diagnostic capabilities of non-expert physicians; diagnostic capabilities of various biopsy procedures; and EUS findings contributing to diagnosis

—

Countries

Japan

Contacts

Public ContactKuwahara Takamichi

Nagoya Graduate School of Medicine Department of Gastroenterology and Hepatology

kuwa_tak@aichi-cc.jp052-744-2172

Outcome results