Artificial intelligence model could classify colorectal polyps as accurately as pathologists

Written by Louis Gautier, Future Science Group

An artificial intelligence (AI) model has been trained to identify colorectal polyps and has achieved the same degree of accuracy as practising pathologists. Researchers also discovered that AI trained on data from a single institution performed as well as models trained on data from multiple institutions. Researchers from the Norris Cotton Cancer Center, based at the Dartmouth-Hitchcock Medical Center (NH, USA), have trained an AI model to accurately classify polyps excised during colonoscopies. Their model can distinguish the four major types of colorectal polyp as accurately as practising clinicians, paving the way for significant improvements to colorectal cancer screening. The...

To view this content, please register now for access

It's completely free