Improving Capsule Endoscopy Diagnosis Using Artificial Intelligence to Reduce Reading Time and Enhance Accuracy

What this Trial Is About

Researchers are studying how artificial intelligence AI can improve capsule endoscopy, a minimally invasive procedure that captures many images of the digestive tract using a small camera capsule. The study focuses on developing and evaluating AI tools to enhance diagnostic accuracy and reduce the time doctors need to review these images. This research aims to address challenges like long reading times and potential errors due to reader fatigue. Participants who have a capsule endoscopy as part of their regular medical care are invited to contribute de-identified images and videos. These data will be used to create databases that train and test AI technologies designed to detect anatomical landmarks, identify abnormalities, and assess cleanliness of the intestines. The AI components will be developed and evaluated individually and combined over several years. During the study, clinical information such as age, gender, exam indication, and type of capsule will be collected. Expert reviewers will classify images to guide AI training. Researchers will assess the AIs diagnostic accuracy and reading efficiency over up to five years. Participants simply provide data from their routine exams, with no additional intervention or treatment involved.

Development of New Diagnostic Tools in Capsule Endoscopy