Development and evaluation of a deep learning model to improve the usability of polyp detection systems during interventions

Since the introduction of commercially available AI‐systems for colorectal polyp detection, theuse of these promising systems in daily practice is increasing. The great potential of AI‐systems iscurrently in the field of diagnostics, as CADe systems support the examiner in real time and withhigh sensitivity.Our novel AI system detects inserted instruments with high sensitivity and specificity. Therefore,the system can capture the time frame of an endoscopic intervention with high accuracy. Thiswould enable the suppression of the CADe signal for the duration of the intervention to focusthe investigator’s concentration on the intervention. The suppression relates not only to falsepositive detections but instead to all CADe detections during an intervention that do not add valueto the endoscopic image. The requirements for such a tool detection system are very high, assuppression of the CADe signal outside of an intervention (false positive instrument detection)may increase the risk of missing other visible polyps. Our study shows that our new AI systemachieves a very high specificity, which is sufficient for this purpose. To obtain this high specificity,our system was trained with a large number of images from multiple centers using differentendoscopy processors. The number of training images we used is comparable to the number usedin development of other CADe systems. In addition, the optimized algorithm presents only a shortdelay of ۴۶۷ ms, that allows for the real‐time use in combination with a CADe system.Since the sensitivity of our AI system is in a high range, the instruments introduced were missedin only a few frames during an intervention. This applies in particular to the insertion and removalof an instrument where only a small portion of it is visible at the edge of the endoscopic view.Once the instrument is in the normal working position, it is quickly and reliably detected by the AIsystem. Thus, the crucial part of the intervention is captured by our instrument detection system.However, a problem with instrument detection arises when an instrument is pressed so firmly intothe mucosa that it is barely visible. In this situation, the instrument recognition works accordinglyworse. Nevertheless, our video analysis showed that the new AI system significantly reduced thenumber of false‐positive CADe detections during an endoscopic intervention. While many publicationson AI systems only use short, specially selected video sequences in the evaluation phase,our system was tested on full‐length colonoscopy, which brings the results much closer to the realexamination situation.Interestingly, the commercially available CADe system seems to generate more detections whena snare is used in comparison to a grasper. There might be different explanations for this phenomenon.Artificial intelligence (AI) for colonic polyp detection is the most important application ofthis new technology in gastrointestinal endoscopy to date. Efficiency and functionality of thesecomputer‐aided detection (CADe) systems have been demonstrated in several randomized trials.However, CADe systems also show many false positive (FP) detections. These false markings canaffect the examiner’s concentration. If a false detection occurs in addition to a relevant finding, theexaminer’s attention may be distracted, leadin to missed findings in the worst case.Therefore, the aim of the current topic was to develop and evaluate an AI system that reliablydetects introduced instruments in order to disable the CADe system during an intervention andavoid distracting detections.