Comparing Real and Artificial Colon Polyp Images for Physician Training in Paris Classification

What this Trial Is About

Training physicians to recognize early signs of colorectal cancer is essential for improving detection of precancerous colon polyps. This clinical trial compares the effectiveness of training doctors using either real colon polyp images or artificial images generated by artificial intelligence AI. The artificial images are designed to be realistic, compliant with data protection, and can represent different precancerous stages in the same location. Participants will be randomly assigned to one of two groups one group will train using the Lutetia Training Platform with real colon polyp images, and the other group will train using the same platform but with AI-generated artificial images. Both groups will practice classifying polyps according to the Paris classification system, which categorizes polyp types based on visual appearance. Doctors taking part will be evaluated over a 9-month period to measure how accurately they classify polyps using the Paris classification. Researchers will also assess factors such as the range of misclassifications, how prior endoscopy experience and time spent on the training impact accuracy, and how regularly participants use the Paris classification in their practice. This study aims to determine which training method better supports physicians in correctly identifying colon polyps.

Training Physicians to Differentiate the Paris Classification Using Artificial Colon Polyp Images