Dr. William E Karnes MD, AGAF

I am a clinical and translational gastroenterologist/scientist with a primary interest in colorectal cancer prevention. After several years as an RO1-funded researcher and Director of Research in the Colorectal Neoplasia Clinic at the Mayo Clinic, I transitioned to private practice for 10 years and was subsequently recruited to UCI in 2012 to spearhead efforts to reduce colorectal cancer burden in Orange County. As part of these efforts, I established a browser-based Colonoscopy Quality Database to promote high-quality colonoscopy exams among Orange County providers. This database now contains over 12K colonoscopies and 28K polyp images with pathology, providing unique opportunities for research. An area of particular interest is the application of artificial intelligence to the challenges of improving and reporting quality in colonoscopy. Toward this aim, I co-founded Docbot, a UCI Applied Innovations Company to develop neural networks that can accurately record quality measures and assist discovery and characterization of precancerous lesions during live colonoscopy. Utilizing our large UCI database we have developed several highly accurate deep learning algorithms, including polyp detection/localization, optical pathology, IBD mucosal severity scoring, and measures of cecal intubation, preparation quality, and withdrawal time. These algorithms can simultaneously process and display outputs during live colonoscopy and are under clinical validation.