Study Suggests Computer-Aided Quality Assurance More Important Than Computer-Aided Detection in Polyp Detection
Douglas K. Rex, MD, MASGE, reviewing Yao L, et al. Endoscopy 2021 Nov 25.
Artificial intelligence (AI) that highlights lesions seen at endoscopy is called computer-aided detection (CADe) or detection software. Computer-aided quality (CAQ) programs assess the completeness of mucosal inspection. In a 4-arm study, standard colonoscopy was compared with CADe, CAQ, and a combination of CADe and CAQ. The CAQ program measured the frame speed exposed per unit time and sent a signal if the examiner was going too fast.
In a study performed in China involving 1076 persons randomized to the 4 groups, adenoma detection rates with the control, CADe, CAQ, and combination CADe-CAQ groups were 14.76%, 21.27%, 24.54%, and 30.6%, respectively. The combination of CADe and CAQ was superior to CADe but not CAQ.
COMMENTIf these data are reproduced, they suggest that CAQ is more important than CADe for detection during colonoscopy and that the combination of CADe and CAQ is ideal. I’m sure we’ll see many more studies on this issue. AI in colonoscopy will likely need to include CAQ to truly reduce the enormous operator dependence of the procedure.
Note to readers: At the time we reviewed this paper, its publisher noted that it was not in final form and that subsequent changes might be made.
CITATION(S)
Yao L, Zhang L, Liu J, et al. An artificial intelligence-based quality improvement system significantly improved the efficacy of computer-aided detection system in colonoscopy: a four group parallel study. Endoscopy 2021 Nov 25. (Epub ahead of print) (https://doi.org/10.1055/a-1706-6174)
