An artificial intelligence-based quality improvement system significantly improved the efficacy of computer-aided detection system in colonoscopy: A four group parallel study.

Background and study aims: Tandem colonoscopy studies have found that about one in five adenomas are missed at colonoscopy. It is still debatable whether the combination of a computer-aided detection (CADe) system for colorectal polyp detection with a computer-aided quality improvement (CAQ) system for real-time withdrawal speed monitoring may result in additional benefits in the task of adenoma detection or if the synergetic effect may be harmed due to excessive visual burden resulting from the information overload. This study aims to evaluate the interaction effect on improving the adenoma detection rate (ADR). Patients and methods: This is a single-center, randomized, four-group parallel controlled study, performed in Renmin Hospital of Wuhan University. Between July 1, 2020 and Oct 15, 2020, 1076 participants were randomly allocated into four treatment groups [control: 271, CADe: 268, CAQ: 269 and CADe plus CAQ (COMBO): 268]. The primary outcome was the ADR. Results: The average ADR in the control, CADe, CAQ and COMBO groups was 14.76% (95% C.I. 10.54-18.98), 21.27% (95% C.I. 16.37-26.17), 24.54% (95% C.I. 19.39-29.68) and 30.6% (95% C.I. 25.08-36.11), respectively. The ADR was higher in the COMBO group compared with the CADe group but not compared with the CAQ group (21.27% VS 30.6%, P=0.024, OR 1.284, 95%C.I. 1.033-1.596; 24.54%vs. 30.6%, P = 0.213, OR = 1.309, 95% C.I. 0.857-2, respectively). Conclusions: CAQ significantly improved the efficacy of CADe in a four-group parallel controlled study. No significant difference in the ADR or PDR was found between the CAQ and COMBO groups.