ONCO-AICO: An AI-Based Educational Tool for Polyp Detection from Colonoscopy

Despite being the gold standard screening tool for colorectal cancer (CRC), colonoscopies suffer from significant miss rates for polyps and adenomas, influenced by various factors including the experience of colonoscopists. In addition, due to the increasing demand for CRC screening, cost-effective solutions to increase the screening capacity of health care institutions explore the potential role of non-physician colonoscopists, such as nurse practitioners. To increase the detection rates for polyps and adenomas among senior and junior colonoscopists, the ED-funded ONCOSCREEN project proposes integrating artificial intelligence (AI) with existing colonoscopy procedures to educate on polyp detection and mitigate human error. The project introduces the ONCO-AICO tool, a web-based platform, through which trainees can annotate colonoscopy videos, with AI providing real-time feedback and suggestions and Explainable AI (XAI) delivering transparent and reliable support for colonoscopy training. By harnessing AI advancements and fostering collaboration, ONCOSCREEN seeks to revolutionize CRC screening and contribute to better patient outcomes.