文献速递腹部MR

基于MRI深度学习预测直肠癌预后的应用

Application of deep learning on MRI for prognostic prediction in rectal cancer.
PubMed 文献 · Eur Radiol · 2026-01-15

本研究回顾性纳入来自3家医院的458例直肠癌患者,随访至少3年,利用治疗前MRI构建并验证联合模型。采用多实例学习(MIL)整合多个切片预测并构建列线图。联合模型总生存期C指数在训练、验证和外部测试队列分别为0.819、0.822和0.759。

要点

  • 联合模型结合MIL深度学习特征与临床特征,预后预测优于单一临床或影像模型。
  • 治疗前MRI即可进行个体化风险分层,无需术后病理信息。
  • MIL可整合多切片信息,适用于肿瘤异质性明显的直肠癌。

核心结论

  • 联合深度学习特征与临床特征的模型增强了预测稳健性,可用于直肠癌治疗前个体化风险分层。
  • MIL提取的深度学习特征与临床知识互补,联合模型利用这种协同效应提升预后评估效能。

研究设计与样本

  • 回顾性多中心研究,纳入458例直肠癌患者,来自三家医院,随访至少3年。
  • 训练集268例,验证集115例,外部测试集75例;收集临床、病理及影像数据。
  • 使用多实例学习(MIL)整合多个切片预测;构建联合深度学习特征与临床病理参数的列线图。
  • 预测效能采用Harrell's C-index和时间依赖性ROC曲线评估。

关键影像表现/征象

  • 治疗前MRI原始图像直接作为深度学习模型输入,无需人工勾画病灶或提取影像组学特征。

诊断标准与关键数值

  • MIL模型OS的C-index:训练队列0.757,验证队列0.754,测试队列0.741。
  • 临床模型OS的C-index:训练队列0.666,验证队列0.772,测试队列0.756。
  • 联合模型OS的C-index:训练队列0.819,验证队列0.822,测试队列0.759。
  • 联合模型DFS的C-index:训练队列0.768,验证队列0.750,测试队列0.721。

临床应用与阅片要点

  • 阅片时可关注治疗前MRI中与预后相关的整体肿瘤形态学信号特征,而非单一层面。
  • 报告可提示结合深度学习辅助工具进行个体化预后评估,辅助新辅助治疗决策。
  • 注意联合模型包含临床病理参数,并非纯影像模型,解读时需结合患者临床信息。

局限性

  • 回顾性研究设计,存在选择偏倚可能。
  • 外部测试集仅75例,样本量有限。

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: Pretreatment MRI was employed to develop and validate a combined model integrating clinical features with deep learning for rectal cancer.

MATERIALS AND METHODS: We retrospectively collected 458 patients from three hospitals and followed them up for at least 3 years. Clinical, pathological and imaging data were collected. Multi-instance learning (MIL) was used to integrate prediction across multiple slices to improve the performance of the model. To improve predictive performance, a nomogram combining deep learning features and clinicopathologic parameters was constructed. Model performance was assessed using Harrell's C-index and time-dependent ROC curves.

RESULTS: The training set included 268 patients, 115 patients in the validation set and 75 patients in the external test set. For OS, the MIL model achieved a C-index of 0.757 in the training cohort, 0.754 in the validation cohort, and 0.741 in the test cohort, compared to 0.666, 0.772, and 0.756 for the clinical model, respectively. The combined model, which integrates MIL features with clinical features, further improved predictive performance, with C-index values for OS at 0.819, 0.822 and 0.759 and for DFS at 0.768, 0.750 and 0.721 across the training, validation and external test cohorts.

CONCLUSIONS: By leveraging the complementary strengths of clinical and deep learning approaches, the combined model enhances predictive robustness, enabling more accurate and personalized pretreatment risk assessment in rectal cancer.

KEY POINTS: Question Rectal cancer management requires more precise prognostic models to optimize treatment strategies and improve clinical decision-making for individual patients. Findings The combined model leverages synergistic effects between clinical and deep learning features, achieving enhanced prognostic performance and enabling more personalized pretreatment risk stratification. Critical relevance This study demonstrates that MIL extracts deep learning features complementary to clinical knowledge. The combined model leverages this synergy, providing clinicians with a more powerful tool for personalized prognostic assessment.

来源:Eur Radiol,查看原文。版权归原作者所有。

引用

  1. Zhu, Gao, Wu, et al. Application of deep learning on MRI for prognostic prediction in rectal cancer. Eur Radiol. 2026. PMID: 41537781. doi:10.1007/s00330-025-12246-0

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。