AI & softwareResearch findings
Development and Validation of Edge AI Device for Gastric Neoplasm Classification in Real-Time Endoscopy
Original published 2024.12.22 09:00 KSTAdded to Blackdeck 2026.11.02 15:01 KST
What has changed?
A real-time edge AI model for classifying all stages of gastric neoplasms during endoscopy is developed
Key points
- A deep learning model for real-time classification of six gastric lesion categories was developed using edge computing
- Internal validation achieved 93.8% accuracy with inference speed of 2–3 ms on GPU and 5–6 ms on CPU
- Endoscopists reported no delays or interference during clinical procedures
Connect it to your work
The model operates seamlessly during real-world endoscopic procedures without delaying diagnosis or disrupting workflow
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.