학술논문

Diagnostic yield of AI-assisted capsule endoscopy reading: an interim analysis of camparison between experts and trainees.
Document Type
Article
Source
Endoscopy. 2024 Supplement 2, Vol. 56, pS64-S64. 1p.
Subject
*ARTIFICIAL intelligence
*CAPSULE endoscopy
*READING
Language
ISSN
0013-726X
Abstract
This article, published in the journal Endoscopy, presents an interim analysis of a study comparing the diagnostic yield and reading time of experts and trainees in capsule endoscopy (CE) with and without the assistance of artificial intelligence (AI). The study evaluated 50 videos of patients who underwent CE, with one expert reviewing the videos in standard mode and the other in AI mode, while trainees evaluated videos in both modalities. The results showed comparable accuracy between standard and AI-assisted readings for both experts and trainees, with AI-assisted reading significantly reducing reading time. The findings suggest that AI could be a valuable optimization strategy for CE training. [Extracted from the article]