文献速递消化道肿瘤与早癌消化内镜

人工智能辅助上消化道内镜检测肿瘤的效果:随机对照试验的系统评价与荟萃分析

Efficacy of artificial intelligence-assisted upper gastrointestinal endoscopy for neoplasm detection: a systematic review and meta-analysis of randomized controlled trials.
PubMed 文献 · Gastrointest Endosc · 2026-02-10

结果显示AI-EGD在每患者和每病灶的肿瘤检出率均显著更高,尤其对≤10mm病变及低级别上皮内瘤变的检出优势更明显。

要点

  • AI-EGD较传统内镜提高上消化道肿瘤每患者检出率(RR 1.57)和每病灶检出率(RR 1.55)。
  • AI-EGD对≤10mm病变检出优势更明显(RR 2.24),>10mm病变检出率相当。
  • AI-EGD对低级别上皮内瘤变、高级别上皮内瘤变和癌的检出率均更高。

核心结论

  • AI-EGD检测上消化道肿瘤优于传统内镜,尤其对≤10mm病变及低级别上皮内瘤变检出率更高。
  • 需进一步试验评估AI在不同经验内镜医师中的影响。

研究设计与样本

  • 系统检索主要文献数据库,纳入比较AI-EGD与Co-EGD的随机对照试验。
  • 主要结局为每患者肿瘤检出率,次要结局包括每病灶检出率及按病变大小和组织学分类分层。
  • 采用随机效应模型合并风险比(RR)及95%CI。

主要发现

  • 每患者肿瘤检出率:AI-EGD组RR 1.57(95%CI 1.23-2.01)。
  • 每病灶肿瘤检出率:AI-EGD组RR 1.55(95%CI 1.33-2.18)。
  • ≤10mm病变检出率:RR 2.24(95%CI 1.72-2.91);>10mm病变检出率相当。
  • 低级别上皮内瘤变检出率:RR 1.73(95%CI 1.30-2.32)。
  • 高级别上皮内瘤变检出率:RR 1.61(95%CI 1.39-1.87);癌检出率:RR 1.54(95%CI 1.32-1.79)。

临床怎么用

  • AI辅助内镜可提高上消化道肿瘤尤其是小病变和低级别上皮内瘤变的检出,支持临床常规使用。
  • 对于>10mm病变,AI辅助的额外检出获益有限。

局限性

  • 需进一步试验评估AI在不同经验水平内镜医师中的效果。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND AND AIMS: EGD remains the primary diagnostic modality for evaluating lesions of the upper gastrointestinal (UGI) tract. This meta-analysis assessed the diagnostic performance of artificial intelligence-EGD (AI-EGD) compared with conventional EGD (Co-EGD) for the detection of UGI neoplasms. METHODS: A systematic search of major bibliographic databases identified randomized controlled trials (RCTs) comparing AI-EGD and Co-EGD in adults undergoing EGD. The primary outcome was neoplasm detection rate (NDR) per patient. Secondary outcomes included NDR per lesion and stratification by lesion size (<10 and ≥10 mm), as well as NDR by histologic classification: low-grade intraepithelial neoplasia (LGIN), high-grade intraepithelial neoplasia (HGIN), and carcinoma. Risk ratios (RRs) with 95% CIs were pooled using a random-effects model. RESULTS: Eleven RCTs involving 57,512 participants were included. AI-EGD demonstrated higher detection rates per patient (RR, 1.57; 95% CI, 1.23-2.01) and per lesion (RR, 1.55; 95% CI, 1.33-2.18). The benefit was more evident for lesions ≤10 mm (RR, 2.24; 95% CI, 1.72-2.91), whereas detection rates for lesions >10 mm were comparable. In addition, AI-EGD achieved higher detection rates across histologic subtypes, including LGIN (RR, 1.73; 95% CI, 1.30-2.32), HGIN (RR, 1.61; 95% CI, 1.39-1.87), and carcinoma (RR, 1.54; 95% CI, 1.32-1.79). CONCLUSIONS: AI-EGD demonstrated superior performance compared with Co-EGD in detecting UGI neoplasms, particularly for lesions ≥10 mm. The greater detection rate of LGIN highlights the potential clinical value of AI support. Further trials are warranted to assess the impact of AI across varying levels of endoscopist experience.

来源:Gastrointest Endosc,查看原文。版权归原作者所有。

引用

  1. Al Hayek M, Barberio B, Al Hayek O, et al. Efficacy of artificial intelligence-assisted upper gastrointestinal endoscopy for neoplasm detection: a systematic review and meta-analysis of randomized controlled trials. Gastrointest Endosc. 2026. PMID: 41672109. doi:10.1016/j.gie.2026.02.005

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。