AI trumps other imaging techniques for adenoma detection: review

Artificial intelligence (AI) algorithms improve detection of colorectal adenomas more than other modern endoscopic imaging techniques, researchers have found.
The Italian-led systematic review and network meta-analysis shows AI is superior to the use of chromoendoscopy and mucosal visualisation systems, and all three trump the control technique, high definition (HD) white-light endoscopy.
The study included 50 randomised controlled trials (RCTs), involving more than 34,000 people undergoing a full colonoscopy, with two of the trials being Australian, the authors reported in the Lancet Gastroenterology and Hepatology.
Analysis showed that the detection rate of adenomas was 7.4% higher when AI or computer-aided techniques were used to alert the endoscopist to potential lesions compared with the control.