Detection of duodenal villous atrophy on endoscopic images using a deep learning algorithm.

BACKGROUND AND AIMS: Celiac disease with its endoscopic manifestation of villous atrophy (VA) is underdiagnosed worldwide. The application of artificial intelligence (AI) for the macroscopic detection of VA at routine EGD may improve diagnostic performance. METHODS: A dataset of 858 endoscopic images of 182 patients with VA and 846 images from 323 patients with normal duodenal mucosa was collected and used to train a ResNet18 deep learning model to detect VA. An external dataset was used to test the algorithm, in addition to 6 fellows and 4 board-certified gastroenterologists. Fellows could consult the AI algorithm's result during the test. From their consultation distribution, a stratification of test images into "easy" and "difficult" was performed and used for classified performance measurement. RESULTS: External validation of the AI algorithm yielded values of 90%, 76%, and 84% for sensitivity, specificity, and accuracy, respectively. Fellows scored corresponding values of 63%, 72%, and 67% and experts scored 72%, 69%, and 71%, respectively. AI consultation significantly improved all trainee performance statistics. Although fellows and experts showed significantly lower performance for difficult images, the performance of the AI algorithm was stable. CONCLUSIONS: In this study, an AI algorithm outperformed endoscopy fellows and experts in the detection of VA on endoscopic still images. AI decision support significantly improved the performance of nonexpert endoscopists. The stable performance on difficult images suggests a further positive add-on effect in challenging cases.

Endoscopic Treatment of Zenker's Diverticulum: Comparable Treatment Outcomes in Treatment-Naïve and Pretreated Patients.

BACKGROUND AND AIMS: Flexible endoscopic treatment plays an important role in the treatment of Zenker's diverticulum (ZD). This study analyzes long-term symptom control and the rate of adverse events in treatment-naïve patients and patients with recurrence, using the stag beetle knife junior (sb knife jr). METHODS: From August 2013 to May 2019, 100 patients with symptomatic ZD were treated with flexible endoscopy using the sb knife jr. Before treatment, as well as 1 and 6 months afterwards, symptoms were obtained by a nine-point questionnaire, with symptoms weighted from 0 to 4. RESULTS: Overall, 126 interventions were performed. The median follow-up period was 41 months (range 7-74). For the three most frequent symptoms, regurgitation, dysphagia, and dry cough, a significant reduction of the mean score could be achieved, from 2.85/3.45/2.85 before the initial treatment to 0.56/1.09/0.98 6 months later. 17 patients were retreated because of recurrence. Out of these, 12 patients underwent a 2nd, 4 patients a 3rd, and 1 patient a 4th session, respectively. The mean dysphagia score for successfully treated patients could be reduced from initially 2.34 to 0.49/0.33/0.67 after the 1st/2nd/3rd session, the frequency of dysphagia from 3.45 to 0.92/1.00/1.33, and the score for regurgitations from 2.85 to 0.35/1.00/0.67. In first-line treatment, as well as in retreatment, no severe adverse event occurred. CONCLUSION: Patients with ZD can be treated safely and effectively with the sb knife jr. Retreatment leads to equal symptom relief as compared to a successful first-line treatment and is not associated with a higher rate of adverse events.