PURPOSE: Aim of the study was the evaluation of high-definition white light (HDWL) i-Scan endoscopy for diagnostic prediction of histology, inter- and intraobserver agreement for colorectal polyps. METHODS: In this prospective single-center cohort study, a skilled endoscopist, with HDWL i-Scan, predicted on real-time the histology of 150 polyps from 78 patients undergoing colonoscopy. Subsequently, four endoscopists, not involved in the collection of cases, predicted the polyps' histology (neoplastic vs nonneoplastic) and repeated the assessment after 6 months. RESULTS: Real-time assessment of the lesions predicted a correct histology in 138/150 cases with a sensitivity, specificity, and accuracy of 95, 82, and 92 %, respectively. Sensitivity, specificity, and accuracy of post-acquisition evaluation were 87.7, 61.7, and 82.2 %, respectively. The κ values of inter- and intraobserver agreement of prediction of histology were 0.462 (0.373-0.537) and 0.657 (0.523-0.917). Among the subset of polyps (n = 82) with good/excellent quality image, sensitivity and accuracy of HDWL i-Scan improved (97 %, p = 0.003 and 94 %, p = 0.002, respectively) as well as κ value of interobserver agreement (0.699). CONCLUSIONS: HDWL i-Scan technology helps for characterization of polyps of the colon with good accuracy even if it cannot replace, at the moment, the histopathological examination. Reproducibility among operators is supported by a moderate substantial interobserver and intraobserver agreement.
PURPOSE: Aim of the study was the evaluation of high-definition white light (HDWL) i-Scan endoscopy for diagnostic prediction of histology, inter- and intraobserver agreement for colorectal polyps. METHODS: In this prospective single-center cohort study, a skilled endoscopist, with HDWL i-Scan, predicted on real-time the histology of 150 polyps from 78 patients undergoing colonoscopy. Subsequently, four endoscopists, not involved in the collection of cases, predicted the polyps' histology (neoplastic vs nonneoplastic) and repeated the assessment after 6 months. RESULTS: Real-time assessment of the lesions predicted a correct histology in 138/150 cases with a sensitivity, specificity, and accuracy of 95, 82, and 92 %, respectively. Sensitivity, specificity, and accuracy of post-acquisition evaluation were 87.7, 61.7, and 82.2 %, respectively. The κ values of inter- and intraobserver agreement of prediction of histology were 0.462 (0.373-0.537) and 0.657 (0.523-0.917). Among the subset of polyps (n = 82) with good/excellent quality image, sensitivity and accuracy of HDWL i-Scan improved (97 %, p = 0.003 and 94 %, p = 0.002, respectively) as well as κ value of interobserver agreement (0.699). CONCLUSIONS: HDWL i-Scan technology helps for characterization of polyps of the colon with good accuracy even if it cannot replace, at the moment, the histopathological examination. Reproducibility among operators is supported by a moderate substantial interobserver and intraobserver agreement.
Authors: J Pohl; E Lotterer; C Balzer; M Sackmann; K-D Schmidt; L Gossner; C Schaab; T Frieling; M Medve; G Mayer; M Nguyen-Tat; C Ell Journal: Gut Date: 2008-10-06 Impact factor: 23.059
Authors: Mariëlle W E Bouwens; Rogier de Ridder; Ad A M Masclee; Ann Driessen; Robert G Riedl; Bjorn Winkens; Silvia Sanduleanu Journal: World J Gastroenterol Date: 2013-07-21 Impact factor: 5.742
Authors: Dale J Waterhouse; Catherine R M Fitzpatrick; Brian W Pogue; James P B O'Connor; Sarah E Bohndiek Journal: Nat Biomed Eng Date: 2019-04-29 Impact factor: 29.234
Authors: Vinay Sehgal; Avi Rosenfeld; David G Graham; Gideon Lipman; Raf Bisschops; Krish Ragunath; Manuel Rodriguez-Justo; Marco Novelli; Matthew R Banks; Rehan J Haidry; Laurence B Lovat Journal: Gastroenterol Res Pract Date: 2018-08-29 Impact factor: 2.260