Search

Background: We have developed the computer-aided detection (CADe) system using an original deep learning algorithm based on a convolutional neural network for assisting endoscopists in detecting colorectal lesions during colonoscopy. The aim of this study was to clarify whether adenoma miss rate (AMR) could be reduced with CADe assistance during screening and surveillance colonoscopy.Methods: This study was a multicenter randomized controlled trial. Patients aged 40 to 80 years who were referred for colorectal screening or surveillance at four sites in Japan were randomly assigned at a 1:1 ratio to either the "standard colonoscopy (SC)-first group" or the "CADe-first group" to undergo a back-to-back tandem procedure. Tandem colonoscopies were performed on the same day for each participant by the same endoscopist in a preassigned order. All polyps detected in each pass were histopathologically diagnosed after biopsy or resection.Results: A total of 358 patients were enrolled and 179 patients were assigned to the SC-first group or CADe-first group. The AMR of the CADe-first group was significantly lower than that of the SC-first group (13.8% vs. 36.7%, P < 0.0001). Similar results were observed for the polyp miss rate (14.2% vs. 40.6%, P < 0.0001) and sessile serrated lesion miss rate (13.0% vs. 38.5%, P = 0.03). The adenoma detection rate of CADe-assisted colonoscopy was 64.5%, which was significantly higher than that of standard colonoscopy (53.6%; P = 0.036).Conclusion: Our study results first showed a reduction in the AMR when assisting with CADe based on deep learning in a multicenter randomized controlled trial. [ABSTRACT FROM AUTHOR]

Objectives: Antithrombotic therapy is a well-known independent risk factor for bleeding after endoscopic submucosal dissection (ESD) of early gastric cancer (EGC). A novel method of ulcer base closure using an endoloop and endoclips has been reported. This study aimed to evaluate the effectiveness of endoscopic closure using an endoloop and endoclips in preventing post-ESD bleeding in patients undergoing gastric ESD on antithrombotic therapy.Methods: This was a single center, retrospective study. Patients on antithrombotic therapy who underwent gastric ESD were divided into two groups, the closure group and the non-closure group. We analyzed procedural outcomes, post-ESD bleeding rate and factors associated with post-ESD bleeding.Results: Among 400 ESDs with EGCs in 311 patients, 131 ESDs in 110 patients were in the closure group, and 269 ESDs in 217 patients were in the non-closure group (16 patients were overlapped in both groups). Post-ESD bleeding rate was 11.5% (15/131) in the closure group, and 11.9% (32/269) in the non-closure group (p = 0.89). Total sustained closure rate during second look endoscopy was 47.8% (33/69). Post-ESD bleeding rate tended to be lower in the closure group than in the non-closure group for lesions located in the greater curvature (3.6% vs. 11.1%, p = 0.11). In addition, sustained closure rate was significantly higher in the greater curvature than in the lesser curvature (72.0% vs. 34.1%, p < 0.01). Multivariate analysis revealed resection size > 40 mm and heparin bridge were the independent risk factor for post-ESD bleeding.Conclusion: Ulcer base closure using endoloop and endoclips did not prevent post-ESD bleeding in patients on antithrombotic therapy. [ABSTRACT FROM AUTHOR]

Background: Salvage endoscopic resection (ER) has been reported to be effective for patients with local failure of esophageal squamous cell carcinoma (ESCC) after definitive chemoradiotherapy (dCRT). This study aimed to evaluate the long-term outcomes of salvage ER for patients with local failure of ESCC and to identify risk factors associated with disease recurrence after salvage ER. Methods: This study included 45 patients undergoing salvage ER after dCRT during 2000 to 2017. After ER, all patients were required to undergo surveillance esophagogastroduodenoscopy (EGD) once or twice every year, and a computed tomography (CT) examination was repeated every 3 to 6 months. We assessed short-term outcomes and long-term outcomes. Results: Of the 45 patients in this study, the baseline clinical T stage before dCRT was T1 in 80%, 66% of the patients did not have nodal metastasis. The median time from CRT to the detection of local failure was 11 months (range 2–130 months). The en-bloc resection rate was 46%, and the R0 resection rate was 38%, respectively. Stricture occurred after salvage ER for one case, while adverse events such as bleeding or perforation and ER-related death did not occur. After a median observation period of 57 months, recurrence free survival at 3 years was 58%, overall survival was 72%, and disease specific survival was 81%. In multivariate analysis, clinical N stage before CRT was the only independent risk factor of recurrence after salvage ER (p = 0.04). Conclusions: Salvage ER might be effective local treatment in patients with local failure after dCRT. For the patients with clinical N stage, frequent surveillance should be performed. [ABSTRACT FROM AUTHOR]

Introduction: Linked color imaging (LCI) and blue laser imaging-bright (BLI-b) improve the visibility of gastrointestinal lesions. In this multicenter study, we compared the effects of LCI and BLI-b on the visibility of flat polyps with visibility scores and color difference (CD) values, including fast-withdrawal and large-monitor observation.Methods: We recorded 120 videos of 40 consecutive flat polyps (2-20 mm), adenoma, and sessile serrated adenoma and polyp (SSA/P), using white light imaging (WLI), BLI-b, and LCI from July 2017 to December 2017. All videos were evaluated by eight endoscopists according to a published polyp visibility score of 4 (excellent) to 1 (poor). Additionally, 1.5 ×faster and 1.7 ×sized videos were evaluated. Moreover, we calculated the CD values for each polyp in three modes.Results: The mean LCI scores (3.1 ± 0.9) were significantly higher than the WLI scores (2.5 ± 1.0, p < 0.001) but not significantly higher than the BLI-b scores (3.0 ± 1.0). The scores of faster videos on LCI (3.0 ± 1.1) were significantly higher than WLI (2.0 ± 1.0, p < 0.001) and BLI-b (2.8 ± 1.1, p = 0.03). The scores of larger-sized videos on LCI were not significantly higher than those of WLI or BLI-b. The CD value of LCI (18.0 ± 7.7) was higher than that of WLI (11.7 ± 7.0, p < 0.001), but was not significantly higher than that of BLI-b (16.6 ± 9.6). The CD value of LCI was significantly higher than that of BLI-b for adenoma, but the CD value of BLI-b was significantly higher than that of LCI for SSA/P.Conclusions: The superiority of LCI to BLI-b was proven for the visibility of adenoma and fast observation. [ABSTRACT FROM AUTHOR]

This article was updated to correct the ordering of the data in Table 6. [ABSTRACT FROM AUTHOR]