Assisted documentation as a new focus for artificial intelligence in endoscopy: the precedent of reliable withdrawal time and image reporting.

This article explores the use of artificial intelligence (AI) in endoscopy to enhance the documentation process. The authors have developed a deep learning-based system that can automatically measure withdrawal time and generate image reports of the cecum and polypectomies. The system was tested on 100 videos and applied in real-time during 10 additional examinations. The results indicate that the AI predictions for withdrawal time were more accurate than the reported times, and the generated image reports were deemed reliable by gastroenterologists. This AI system has the potential to improve standardized colonoscopy documentation and reduce the workload for endoscopists. Further research is needed to evaluate and expand the capabilities of AI in colonoscopy reporting. [Extracted from the article]