Enhancing polyp detection and patient outcomes
Colonoscopy plays a vital role in the early detection and prevention of colorectal cancer, through the removal of adenomatous and serrated colonic polyps, which are the precursors to colon cancer.
Adenoma and serrated polyp detection rates have been shown to correlate with the risk of missed colon cancer1, with higher detection inferring a lower risk of missed lesions. However, even with the skill of experienced gastroenterologists and improved imaging resolution, some polyps may not be easily seen during colonoscopy.
This is where computer-aided detection (CAD) systems come into play, to improve the effectiveness of colonoscopy through increased polyp detection.
CAD utilises algorithms developed by machine based learning based on thousands of images of colonic polyps. It then applies those algorithms to real-time digital images during colonoscopy. By highlighting potential polyps, it serves to boost polyp detection, aiding in the identification of polyps that may have been missed by the naked eye.
The key benefit of Computer-Aided Detection (CAD) Systems
Its ability to assist healthcare professionals in real-time, providing immediate feedback during the procedure. It helps reduce the risk of missing polyps, especially small or flat polyps, or polyps that are difficult to detect due to location or subtle appearance. Some CAD systems can also help differentiate between two main classes of colonic polyp to further boost the confidence of the operator in differentiating polyp type. CAD has been shown to increase polyp detection of even experienced endoscopists.
A recently published study2 found a greater than four percentage absolute improvement in adenoma detection rates versus standard colonoscopy techniques. With the improvement occurring in routine colonoscopies performed by 20 non-academic endoscopists, the findings support the benefits of computer-aided detection (CAD) in everyday practice, according to the researchers.
While CAD systems significantly enhance polyp detection rates, they are not intended to replace the expertise of gastroenterologists. Instead, they serve as an additional tool to improve accuracy and increase the overall efficacy of colonoscopy procedures.
CAD technology is relatively new and tends to be oversensitive, it takes a well-trained and experienced gastroenterologist to asses multiple false positives and also the false negatives when activated. As gastroenterologists, we must still perform a high quality colonoscopy and take time to expose as much of the colonic mucosa as possible during the examination to reduce the risk of missed polyps or cancer.
1.Adenoma detection rate and risk of colorectal cancer and death.Corley DA et alN Engl J Med. 2014 Apr 3;370(14):1298-306. doi: 10.1056/NEJMoa1309086.PMID: 24693890
2.Effect of real-time computer-aided detection of colorectal adenoma in routine colonoscopy : a single-centre randomised controlled trial , Karsenti.D et al. (May 31, 2023), The Lancet Gastroenterology & Hepatology DOI:https://doi.org/10.1016/S2468-1253(23)00104-8