人工智能 (AI) 越来越多地用于改善结肠镜检查期间的腺瘤检测。这项荟萃分析旨在提供对计算机辅助检测 (CADe) 系统及其对关键结肠镜检查质量指标的影响的最新评估。我们从 Embase、PubMed 和 MEDLINE 数据库中检索了从一开始到 2024 年 2 月 15 日的随机对照试验（ RCT）将 CADe 系统与常规无辅助结肠镜检查在结直肠腺瘤检测方面的性能进行比较。选择纳入 28 项 RCT，涉及 23861 名参与者。随机效应荟萃分析显示，ADR 增加 20%（RR 1.20，95% CI 1.14-1.27，p<0.01），AMR 降低 55%（RR 0.45，95% CI 0.37-0.54，p<0.01）。人工智能辅助结肠镜检查。仅涉及内窥镜专家的亚组分析显示出相似的效应大小（RR 1.19，95% CI 1.11-1.27，p <0.001），在不同 CADe 系统和医疗保健环境的分析中也发现了类似的结果。使用 CADe 还显着增加了每次结肠镜检查的腺瘤数量（加权平均差 0.21，95% CI 0.14-0.29，p<0.01），这主要是由于微小病变检出率的增加，而晚期腺瘤检出率没有显着差异。无蒂锯齿状病变的检测（RR 1.10，95% CI 0.93-1.30，p=0.27）和漏检率（RR 0.44，95% CI 0.16-1.19，p=0.11）相似。 AI 辅助结肠镜检查的撤药时间平均延长 0.15 分钟（加权平均差 0.15，95% CI 0.04-0.25，p = 0.01），非肿瘤性切除率增加 39%（RR 1.39，95）。 % CI 1.23-1.57，p<0.001）。AI 辅助结肠镜检查显着改善了腺瘤的检测，但不改善无蒂锯齿状病变，无论内窥镜医师经验、系统类型或医疗保健环境如何。版权所有 © 2024 美国胃肠内窥镜协会。由爱思唯尔公司出版。保留所有权利。

Artificial intelligence (AI) is increasingly used to improve adenoma detection during colonoscopy. This meta-analysis aimed to provide an updated evaluation of computer-aided detection (CADe) systems and their impact on key colonoscopy quality indicators.We searched the Embase, PubMed and MEDLINE databases from inception until February 15, 2024, for randomised control trials (RCTs) comparing the performance CADe systems with routine unassisted colonoscopy in the detection of colorectal adenomas.28 RCTs were selected for inclusion involving 23861 participants. Random-effects meta-analysis demonstrated a 20% increase in ADR (RR 1.20, 95% CI 1.14-1.27, p<0.01) and 55% decrease in AMR (RR 0.45, 95% CI 0.37-0.54, p<0.01) with AI-assisted colonoscopy. Subgroup analyses involving only expert endoscopists demonstrated a similar effect size (RR 1.19, 95% CI 1.11-1.27, p<0.001), with similar findings seen in analysis of differing CADe systems and healthcare settings. CADe use also significantly increased adenomas per colonoscopy (weighted mean difference 0.21, 95% CI 0.14-0.29, p<0.01), primarily due to increased diminutive lesion detection, with no significant difference seen in detection of advanced adenoma. Sessile serrated lesion detection (RR 1.10, 95% CI 0.93-1.30, p=0.27) and miss rates (RR 0.44, 95% CI 0.16-1.19, p=0.11) were similar. There was an average 0.15 minute prolongation of withdrawal time with AI-assisted colonoscopy (weighted mean difference 0.15, 95% CI 0.04-0.25, p = 0.01) and a 39% increase in the rate of non-neoplastic resection (RR 1.39, 95% CI 1.23-1.57, p<0.001).AI-assisted colonoscopy significantly improved adenoma, but not sessile serrated lesion, detection irrespective of endoscopist experience, system type or healthcare setting.Copyright © 2024 American Society for Gastrointestinal Endoscopy. Published by Elsevier Inc. All rights reserved.