Artificial intelligence increases adenoma detection in CRC screening
The addition of real-time computer-aided detection in colonoscopy significantly increased the adenoma detection rate and adenomas detected per colonoscopy in colorectal cancer screening, according to a presentation at ESGE Days.
“[An] expert endoscopist can benefit from the artificial intelligence, increasing the detection,” Alessandro Repici, MD, professor of gastroenterology, director of digestive endoscopy unit at Humanitas Research Hospital in Rozzano, Italy, said during his presentation. “This benefit is prevalent when adhered by the expert endoscopist including an interference between the [computer-aided detection (CAD-e)] and the level of experience of the operator.”
At five European centers, Repici and colleagues assessed 660 patients who underwent screening colonoscopies for CRC, post-polypectomy surveillance, or workup due to positive results from a fecal immunochemical test or signs or symptoms of CRC. Investigators randomly assigned patients to either high-definition colonoscopies with real-time computer-aided detection (CAD-e group) or without (control group). A minimum withdrawal time of 6 minutes was needed. Adenoma detection rate served as the primary outcome. Other outcomes included adenomas detected per colonoscopy and withdrawal time.
The adenoma detection rate in the CAD-e group (53.3%) was significantly higher compared with the control group (44.2%; OR = 1.44; 95% CI, 1.06-1.96). In addition, adenomas detected per colonoscopy was significantly higher in the CAD-e group compared with the controls (1.26; 95% CI,1.14-1.38 vs. 1.04; 95% CI, 0.93-1.15; incident rate ratios [IRR] = 1.21; 95% CI,1.05-1.4).
Researchers observed no difference in withdrawal time (CAD-e, 8.1±1.61 minutes vs. control, 7.9±1.53; P = .06).
“What we found is that the use of the artificial intelligence but not the level of the examiner experience was associated with ADR differences,” Repici said.