Computer-aided detection-assisted colonoscopy: classification and relevance of false positives.

Background and Aims False-positive results (FPs) by computer-aided detection (CADe) hamper colonoscopy efficiency by uselessly extending examination time. Our aim was to develop a classification of the causes and clinical relevance of CADe FPs, and to assess the relative distribution of FPs in real-life setting. Methods In a post-hoc analysis of a randomized trial comparing colonoscopy with and without CADe (NCT: 04079478), we extracted 40 CADe-colonoscopy videos. By modified Delphi process, 4 expert endoscopists identified the main domains for reasons and clinical relevance of FPs. Then, 2 expert endoscopists manually examined each FP activation, and classified it according to the proposed domains. Analysis was limited to the withdrawal phase. Results The 2 main domains for the causes of CADe-FPs were identified as artifacts due to either mucosal wall or bowel content, whereas clinical relevance was defined by the time spent for FP and the FP rate per minute. Mean FPs per colonoscopy was 27.3 ± 13.1, of which 24 ± 12 (88%) and 3.2 ± 2.6 (12%) due to artifacts from bowel wall and bowel content, respectively. Of the 27.3 FP per colonoscopy, 1.6 (5.7%) required an additional exploration time of 4.8 ± 6.2 seconds per FP, ie, 0.7% of the mean withdrawal time (WT). In detail, 15 (24.2%), 33 (53.2%) and 14 (22.6%) FPs were classified as of mild, moderate, or severe clinical relevance. Rate of FP per minute of WT was 2.4 ± 1.2, being higher for FPs due to artifacts from bowel wall than for those from bowel content (2.4 ± 0.6 vs 0.3 ± 0.2, p Conclusions False positives by CADe are primarily due to artifacts from the bowel wall. Despite a high frequency, FPs result in a negligible 1% increase of the total withdrawal time as most of them are immediately discarded by the endoscopists.