文献速递腹部技术与安全CT

基于计算机断层扫描(CT)的影像组学与人工智能(AI)模型诊断肝细胞癌的诊断准确性:系统综述与Meta分析

Diagnostic accuracy of computed tomography (CT)-based radiomics and artificial intelligence (AI) models in hepatocellular carcinoma: a systematic review and meta-analysis.
PubMed 文献 · Clin Radiol · 2025-08-08

本系统综述与Meta分析检索2015至2025年四个主要数据库,纳入29项研究,其中17项提供独立性能指标进入Meta分析,评估基于增强CT的机器学习和影像组学模型诊断肝细胞癌的效能。在独立测试集上,合并敏感度和特异度分别为0.562和0.570;影像组学专用模型(n=3)敏感度和特异度分别为0.660和0.670,敏感度异质性I²=0%;联合模型敏感度和特异度分别为0.368和0.297,异质性I²=98.3%。作者结论认为影像组学模型在独立数据集上诊断准确性更一致,但方法学变异大,需要标准化报告、稳健外部验证和可解释AI技术。

要点

  • 独立测试集上AI/影像组学模型合并敏感度0.562、特异度0.570,诊断准确性有限。
  • 影像组学专用模型(n=3)敏感度0.660、特异度0.670,敏感度异质性I²=0%。
  • 联合模型敏感度0.368、特异度0.297,异质性I²=98.3%,可能因缺乏独立测试集。

核心结论

  • 基于独立数据集评估的影像组学模型对增强CT诊断肝细胞癌表现出更一致的诊断准确性,但方法学变异大,需要标准化报告、稳健外部验证和可解释AI技术以促进临床采用。

研究设计与样本

  • 系统综述与Meta分析,检索2015至2025年四个主要数据库,按PRISMA指南纳入29项研究,采用QUADAS-2评估偏倚风险,其中17项提供独立性能指标进入Meta分析,使用随机效应模型合并敏感度、特异度、准确性和F1评分。

诊断标准与关键数值

  • 独立测试集上模型合并敏感度0.562、特异度0.570;影像组学专用模型(n=3)敏感度0.660、特异度0.670,敏感度异质性I²=0%、τ²≈0;联合模型敏感度0.368、特异度0.297,异质性I²=98.3%。

局限性

  • 影像组学专用模型研究数量有限(n=3),结果应谨慎解释。
  • 联合模型异质性高(I²=98.3%),可能因缺乏独立测试集导致泛化性有限。
  • 方法学变异大,需要标准化报告、稳健外部验证和可解释AI技术。

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)。

英文原文(摘要)· 点击展开

AIM: Hepatocellular carcinoma (HCC) is among the leading causes of cancer-related deaths worldwide. The clinical utility of artificial intelligence (AI) and radiomics models based on contrast-enhanced computed tomography (CECT) for HCC detection remains underexplored. This systematic review and meta-analysis evaluated the diagnostic performance of machine learning and radiomics-based models.

MATERIALS AND METHODS: A comprehensive literature search (2015-2025) across four major databases identified 29 eligible studies according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. The risk of bias was assessed using the QUADAS-2 tool. Of these, 17 studies provided independent performance metrics and were included in the meta-analysis. Pooled estimates of sensitivity, specificity, accuracy, and the F1 score were calculated using a random effects model.

RESULTS: Models evaluated on independent test datasets, the pooled sensitivity and specificity were 0.562 and 0.570, respectively. A subgroup of radiomics-specific models (n=3) demonstrated comparatively higher sensitivity (0.660) and specificity (0.670), with minimal heterogeneity in sensitivity (I2 = 0%, τ2 ≈ 0), suggesting strong internal consistency. However, due to the limited number of studies, these results should be interpreted with caution. Combined models showed lower sensitivity (0.368) and specificity (0.297), with high heterogeneity (I2 = 98.3%), likely due to the absence of independent test sets, which may have led to limited generalisability.

CONCLUSION: Radiomics-based models evaluated on independent datasets exhibit more consistent diagnostic accuracy for HCC detection using CECT. However, methodological variability underscores the need for standardised reporting, robust external validation, and incorporation of explainable AI techniques to enhance clinical adoption.

来源:Clin Radiol,查看原文。版权归原作者所有。

引用

  1. Sweta K, Dkhar W, Kadavigere R, et al. Diagnostic accuracy of computed tomography (CT)-based radiomics and artificial intelligence (AI) models in hepatocellular carcinoma: a systematic review and meta-analysis. Clin Radiol. 2025. PMID: 40907089. doi:10.1016/j.crad.2025.107042

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。