Creation of an automatic diagnosis system for endoscopic images of esophageal disease using artificial intelligence
Not Applicable
- Conditions
- Esophageal diseases (such as esophageal malignant tumors, leiomyomas, neuroendocrine tumors, granuloma, reflux esophagitis, and eosinophilic esophagitis) and normal esophagus
- Registration Number
- JPRN-UMIN000039645
- Lead Sponsor
- Osaka International Cancer Institute Gastrointestinal Oncology
- Brief Summary
We prepared 100 video datasets, including 50 superficial ESCCs, 22 noncancerous lesions, and 28 normal esophagi. The AI system had sensitivity of 85.7% (54 of 63 ESCCs) and specificity of 40%. Initial evaluation by endoscopists conducted with plain video (without AI support) had average sensitivity of 75.0% (47.3 of 63 ESCC) and specificity of 91.4%.
- Detailed Description
Not available
Recruitment & Eligibility
- Status
- Complete: follow-up complete
- Sex
- All
- Target Recruitment
- 21
Inclusion Criteria
Not provided
Exclusion Criteria
When the patient reject to use existing information through an information disclosure document published on the homepage of the facility.
Study & Design
- Study Type
- Observational
- Study Design
- Not specified
- Primary Outcome Measures
Name Time Method
- Secondary Outcome Measures
Name Time Method