AI & softwareResearch findings
Development of a Multidevice CNN for Automatic Detection of Clinically Relevant Lesions in Panendoscopy
Original published 2024.01.01 09:00 KSTAdded to Blackdeck 2026.11.16 04:20 KST
What has changed?
First multidevice CNN model for automatic detection of clinically relevant lesions in panendoscopy
Key points
- Developed a multidevice CNN model to automatically detect clinically relevant lesions during panendoscopy
- Model achieved 88.9% sensitivity and 98.9% specificity, demonstrating high clinical reliability
- Trained on diverse device types including single, double, and motorized spiral enteroscopies to enhance generalizability
Connect it to your work
Trained on 338 endoscopic exams across multiple device types
What remains unproven
This research note is based on an abstract; check the original for test conditions and application scope.
Originals used for this note
Links open the original publisher in a new tab.