文献速递腹部技术与安全MR

肝脏MRI影像组学可重复性与判别力:深度学习重建的影响

Reproducibility and discriminative power of MRI liver radiomics: Impact of deep learning-based image reconstruction.
PubMed 文献 · Eur J Radiol · 2026-08-05

这项回顾性研究纳入95例肝脏恶性病变患者,行1.5T MRI(T2、T2FS与扩散加权成像),比较传统重建与DLIR(AIR Recon DL)对影像组学特征的影响。由两名阅片者分割肿瘤、瘤周与健康肝脏ROI,以ICC评价可重复性,以随机森林和Wilcoxon检验评价组织区分能力。结果显示DLIR未显著改善或降低可重复性,且保留组织区分所需特征,支持其用于影像组学流程。

要点

  • DLIR 未显著改变肝脏 MRI 影像组学特征的可重复性与组织区分能力,可用于临床影像组学流程。
  • 观察者间可重复性整体较低,尤其强度特征,阅片与分割需保持一致。
  • DLIR 使可区分特征比例下降 2-7%,对细微纹理有轻度平滑效应。

核心结论

  • 在肝脏 MRI 中,DLIR 未显著改善或降低影像组学特征的可重复性。
  • DLIR 保留组织区分所需的基本特征,可在影像组学流程中整合而不损害定量影像生物标志物可靠性。

研究设计与样本

  • 回顾性研究,纳入 95 例肝脏恶性病变患者,行 1.5T MRI,序列包括 T2、T2FS 与扩散加权成像。
  • 图像分别采用传统非 DLIR 与 DLIR(AIR Recon DL)重建,由两名阅片者分割肿瘤、瘤周与健康肝脏三个 ROI。
  • 观察者内与观察者间可重复性采用 ICC 评价,区分组织类型的判别能力采用随机森林分类器与 Wilcoxon 检验评估。

诊断标准与关键数值

  • 广义线性混合模型显示,DLIR 与非 DLIR 观察者内可重复特征数量无显著差异,OR 为 0.860,95%CI [0.653; 1.135]。
  • DLIR 与非 DLIR 观察者间可重复特征数量无显著差异,OR 为 1.098,95%CI [0.673; 1.791]。
  • 各重建方式下观察者间可重复性均较低,尤其强度特征。
  • DLIR 下可区分特征比例较非 DLIR 下降 2-7%,提示对细微纹理有轻度平滑效应。

临床应用与阅片要点

  • 临床肝脏 MRI 影像组学流程中可整合 DLIR,因为其对特征可重复性与组织区分能力无显著影响。
  • 报告或研究影像组学结果时,应关注观察者间分割一致性,尤其是强度特征的可重复性较低。
  • DLIR 对细微纹理有轻度平滑效应,基于纹理特征的分析需考虑这一影响。

局限性

  • 本研究为回顾性设计。
  • 本研究仅评价 1.5T MRI 与 AIR Recon DL 一种深度学习重建方法。

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

OBJECTIVES: While Deep Learning-based Image Reconstruction (DLIR) is increasingly used to improve MRI quality, its influence on radiomic stability remains underexplored. This study aimed to assess the impact of DLIR on the reproducibility and discriminative power of radiomic features in liver MRI.

MATERIAL AND METHODS: In this retrospective study, 95 patients with malignant liver lesions underwent 1.5 T MRI (T2, T2FS, and Diffusion). Images were reconstructed using both traditional (non-DLIR) and DLIR (AIR Recon DL) methods. Three regions of interest (ROI) - tumor, peritumoral, and healthy liver - were segmented by two readers. Intra- and inter-observer reproducibility were evaluated using ICC. Discriminative power was assessed via Random Forest classifiers and Wilcoxon tests to distinguish tissue types.

RESULTS: Generalized Linear Mixed Models showed no significant difference in the number of reproducible features between DLIR and non-DLIR for intra-observer (Odds Ratio (OR): 0.860, 95%CI [0.653; 1.135]) or inter-observer (OR: 1.098, 95%CI [0.673; 1.791]) analyses. Inter-observer reproducibility remained low across all reconstructions, particularly for intensity features. Classifier performance for tissue differentiation remained comparable between reconstructions. A slight decrease (2-7%) in the percentage of discriminative features was observed with DLIR, suggesting a mild smoothing effect on subtle textures.

CONCLUSION: DLIR does not significantly improve or degrade the reproducibility of radiomics in liver MRI. However, DLIR preserves the essential signatures required for tissue discrimination. These results support the clinical integration of DLIR in radiomic workflows, confirming that it can be used without compromising the reliability of quantitative imaging biomarkers.

来源:Eur J Radiol,查看原文。版权归原作者所有。

引用

  1. Rey, Rabbani, Chauveau, et al. Reproducibility and discriminative power of MRI liver radiomics: Impact of deep learning-based image reconstruction. Eur J Radiol. 2026. PMID: 42570578. doi:10.1016/j.ejrad.2026.113140

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。