Search

Objectives: Recently, CAD EYE (Fujifilm, Tokyo, Japan), an artificial intelligence for the lesion recognition (CADe) and the optical diagnosis (CADx) of colorectal polyps, was released. We evaluated the function of CADe and CADx of CAD EYE. Methods: In this single-center retrospective study, we examined consecutive polyps ≤ 10 mm detected from March to April 2021 to determine whether CAD EYE could recognize them live with both normal- and high-speed observation using white-light imaging (WLI) and linked-color imaging (LCI). We then examined whether the polyps were neoplastic or hyperplastic live with magnified or non-magnified blue-laser imaging (BLI-LASER) or blue-light imaging (BLI-LED) under CAD EYE, comparing the retrospective evaluations with 5 experts and 5 trainees using still images. All polyps were histopathologically examined. Results: We analyzed 100 polyps (mean size 3.9 ± 2.6 mm; 55 neoplastic and 45 hyperplastic lesions) in 25 patients. Regarding CADe, the respective detection rates of CAD EYE with normal- and high-speed observation were 85.0% and 67.0% for WLI (p = 0.002) and 89.0% and 75.0% for LCI (p = 0.009). Regarding CADx for differentiating neoplastic and hyperplastic lesions, the diagnostic accuracy values of CAD EYE with non-magnified and magnified BLI-LASER/LED were 88.8% and 87.8%. Regarding magnified BLI-LASER/LED, the diagnostic accuracy value of CAD EYE was not significantly different from that of experts (92.0%, p = 0.17), but that of trainees (79.0%, p = 0.04). We also found no significant differences in CADe or CADx between LED (53 lesions) and LASER (47 lesions). Conclusions: CAD EYE was a helpful tool for CADe and CADx in clinical practice. [ABSTRACT FROM AUTHOR]

Background: Local recurrence after cold snare polypectomy (CSP) of colorectal polyps has not been well analyzed. In this study, we analyzed the characteristics of local recurrence.Methods: We retrospectively reviewed consecutive lesions resected by CSP from 2014 to 2016 and lesions that were followed up at ≥ 10 months after CSP, were analyzed. Our indication for CSP was a benign tumor of < 15 mm in size. We analyzed local recurrence and its risk factors using multivariate analyses. In addition, we compared lesions of ≥ 10 mm and < 10 mm. Moreover, therapeutic methods for recurrence were analyzed.Results: Finally, we analyzed 554 cases out of 820 consecutive cases. The mean polyp size was 5.3 ± 2.8 mm and the en bloc resection and histopathological complete resection rates were 99.3% and 70.2%, respectively. The overall recurrence rate was 1.9% (mean follow-up period: 13.0 ± 4.0 months). Significant differences were observed between 11 recurrent lesions and 543 lesions without recurrence regarding polyp size (8.0 ± 3.7 mm vs. 5.2 ± 2.7 mm, p < 0.001), rate of sessile-serrated polyp (27.3% vs. 3.0%, p < 0.001), and histopathological positive margin (45.4% vs. 3.7%, p = 0.019). Multivariate analyses showed that a histopathological positive margin was the only significant risk factor for local recurrence (OR 16.600, 95% CI 3.707-74.331, p < 0.001). Regarding the comparison between 74 lesions of ≥ 10 mm and 480 lesions of < 10 mm, significant differences were observed in the en bloc resection rate (93.2% vs. 100%, p < 0.001), high-grade dysplasia rate (8.1% vs. 0.8%, p < 0.001), and histopathological complete resection rate (54.0% vs. 72.7%, p = 0.001). The recurrence rates of these two groups were 5.4% and 1.4%, respectively (p = 0.069). All recurrent cases could be resected with repeat CSP.Conclusions: The local recurrence rate after CSP for lesions of < 10 mm was 1.4%. CSP was not recommended for lesions of ≥ 10 mm due to high rates of recurrence and malignancy. [ABSTRACT FROM AUTHOR]

Objectives: An endoscope with a light-emitting diode (LED) light source which has a 2-mm close-distance observation function without magnification, has been marketed, enabling linked color imaging (LCI) and blue laser imaging (BLI) for tumor detection and characterization. We analyzed the efficacy of a LED endoscope compared to a LASER endoscope. Methods: We retrospectively reviewed 272 lesions observed using the LED endoscopic system (Fujifilm Co., Tokyo, Japan) from May 2018 to September 2019. The Japanese NBI Classification was used for tumor characterization. We analyzed the diagnostic accuracy and confidence level. Sixty-one lesions observed with both the LED and magnified LASER endoscopes were also analyzed to compare the diagnostic accuracy. Regarding the tumor detectability, we calculated color difference values (CDVs) and brightness values (BVs) of white-light imaging, BLI, and LCI modes between the two endoscopes for each tumor. Results: The mean polyp size was 9.2 ± 11.3 mm. Histology showed 71 sessile serrated lesions, 193 adenoma and high-grade dysplasias, and 8 T1 cancers. The diagnostic accuracy of tumors ≥ 10 and < 10 mm was 72.0% and 92.9% (p < 0.001), respectively and the high confidence rate was 93.8%. The diagnostic accuracy of LED (77.0%) was a little higher than that of LASER without magnification (65.6%, p = 0.16) but was not inferior to that of LASER with magnification (82.0%, p = 0.50). The respective CDVs of LED and LASER endoscopes were 20.6 ± 11.2 and 21.6 ± 11.2 for LCI (p = 0.30), and the respective BVs were 210.0 ± 24.2 and 175.9 ± 21.1 (p < 0.001). Conclusions: A LED endoscope with close-distance observation improved tumor detection and characterization due to high brightness. [ABSTRACT FROM AUTHOR]