The Learning Curve Of Confocal Laser Endomicroscopy For Prediction Of Gastric Polys

Background and aimsConfocal laser endomicroscopy (CLE) is an emerging tool for in vivo imaging of the GI tract that requires the endoscopist to interpret microscopic images. The learning curve for interpretation of CLE images is unknown. This study aims to examine the learning curve of CLE in differentiating hyperplastic and adonomatous gastric polyps and to evaluate the interobserver agreement in diagnosing gastric polyps.MethodsPatients with histologically diagnosed gastric hyperplastic polyp or adenoma identified during CLE examinations were recruited from Qilu Hospital of Shandong University. Four confocal images (2superficial images and2deeper images) and1conventional white-light endoscopic images were selected from each of the95lesions in this study by an experienced endomicroscopist. Before evaluation, an educational set including The purpose of this study, basic principles of CLE, A training set of15images with known histology(not included in the evaluation) was firstly reviewed to standardize image interpretation, followed by blinded review of80unknown images. Selected images of the80gastric polyps were divided into4learning stages randomly. Each stage includes20gastric polyps, Selected images were evaluated by4 experienced CLE investigators and4nonexperienced ones independently. Each study stage was assessed using the QILU classification systems at a1-week interval.ResultsFor the NEE group/The accuracy of4part was0.50、0.70.0.90.0.87; and the sensitivity was0,65、0.80.0.94.0.88; and the specificity was0.67、0.80、0.93、0.87.For the LEE group/The accuracy was0.60.0.75.0.95.0.88; and the sensitivity was0.67、0.83、0.97、0.92; the specificity was0.73、0.87、0.95.0.90.Interobserver agreement was substantial for each part (κ=0.44.0.66.0.84.0.81respectively) For the NEE group; and (κ=0.51.0.69、0.87、0.86respectively) For the LEE group; With increasing experience,Whether NEE or LEE group,the accuracy. specificity and sensitivity increased. There was no significant difference in observation times in relation to the level of assessors’expertise.ConclusionsThe QiLu classification systems for the prediction of gastricl hyperplastic polyp or adenoma have a high accuracy, sensitivity and specificity. Accurate interpretation of CLE images for predicting neoplastic lesions can be learned rapidly by a wide range of GI specialists. which indicates a broader clinical application of these CLE diagnostic system for gastric polyps.