文献速递腹部泌尿生殖技术与安全CT

双能CT碘图深度学习预测胃癌PD-L1表达

Deep Learning Analysis Based on Dual-energy CT-Derived Iodine Map for Predicting PD-L1 Expression in Gastric Cancer: A Multicenter Study.
PubMed 文献 · Acad Radiol · 2026-01-16

本研究基于267例接受胃切除及术前双能CT的多中心胃癌患者,从碘图肿瘤感兴趣区提取深度学习特征并构建特征标签模型。结果显示该模型在训练、内部验证及外部验证队列的AUC分别为0.854、0.836、0.818,优于临床模型,可用于无创预测胃癌PD-L1表达水平。

要点

  • 碘图深度学习特征模型可无创预测胃癌PD-L1表达,有助筛选免疫治疗候选者。
  • 该模型在外部验证队列中AUC达0.818,优于单纯临床模型。
  • SHAP与Grad-CAM提供了模型决策的可视化解释。

核心结论

  • 基于碘图的深度学习模型被证实是一种有价值、可靠且可解释的工具,可用于无创预测胃癌PD-L1表达水平。

研究设计与样本

  • 多中心前瞻性纳入267例接受胃切除术及术前双能CT的胃癌患者。
  • 患者分为训练队列143例、内部验证队列60例、外部验证队列64例。
  • 采用50层残差网络从碘图上肿瘤感兴趣区体积中提取深度学习特征。
  • 通过机器学习构建深度学习特征标签模型(DFSigM)。
  • 采用多变量逻辑回归筛选PD-L1相关临床特征,并构建临床模型和深度学习-临床融合模型。

关键影像表现/征象

  • 深度学习特征标签模型与融合模型在训练队列AUC分别为0.854和0.858。
  • 深度学习特征标签模型在内部验证队列AUC为0.836,融合模型为0.828。
  • 深度学习特征标签模型在外部验证队列AUC为0.818,融合模型为0.833。

诊断标准与关键数值

  • 经特征筛选后确定12个关键深度学习特征用于构建DFSigM模型。
  • DFSigM在训练队列AUC为0.854,内部验证队列AUC为0.836,外部验证队列AUC为0.818。
  • 临床模型在训练队列AUC为0.785,内部验证队列AUC为0.720,外部验证队列AUC为0.695。
  • DFSigM在较宽阈值概率范围内提供了较高的净临床获益。
  • DFSigM预测概率与实际概率之间具有良好的校准一致性。

临床应用与阅片要点

  • 双能CT碘图可作为胃癌PD-L1表达无创评估的依据,辅助免疫治疗决策。
  • 碘图深度学习特征模型在外部验证中表现稳定,提示具有潜在的泛化能力。
  • SHAP与Grad-CAM可视化有助于理解模型决策过程,增加临床应用可信度。

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

RATIONALE AND OBJECTIVES: PD-L1 expression is a critical biomarker in guiding immunotherapy for gastric cancer (GC). This study aims to investigate the value of deep learning analysis based on dual-energy CT-derived iodine map for predicting the level of PD-L1 expression in GC.

METHODS: A total of 267 GC patients who underwent gastrectomy and preoperative dual-energy CT from multiple centers were prospectively enrolled and categorized into training (TC, n=143), internal validation (IVC, n=60), and external validation cohort (EVC, n=64). A 50-layer Residual Network was used to extract deep learning (DL) features from tumor volumes of interest on the iodine map. Machine learning was employed to develop the DL feature signature model (DFSigM). Multivariable logistic regression was used to screen PD-L1-related clinical characteristics, then a clinical model and a DL-clinical fusion model were also built. Model performance was evaluated based on discrimination, calibration, and clinical utility. Model interpretability was achieved through SHAP and Grad-CAM.

RESULTS: Following feature selection, 12 key DL features were identified and utilized to construct DFSigM. DFSigM achieved AUC values of 0.854 in TC, 0.836 in IVC, and 0.818 in EVC, outperforming the clinical model (AUCs of 0.785, 0.720, and 0.695), while comparable to the fusion model (AUCs of 0.858, 0.828, and 0.833). DFSigM provided a high net clinical benefit across a wide range of threshold probabilities, and also demonstrated good agreement between the predicted and actual probabilities. SHAP and Grad-CAM visualized the decision-making process of the model.

CONCLUSION: A deep learning model based on iodine map has been proven to be a valuable, reliable, and interpretable tool for non-invasive prediction of PD-L1 expression in GC.

来源:Acad Radiol,查看原文。版权归原作者所有。

引用

  1. Chen, Zhao, Tian, et al. Deep Learning Analysis Based on Dual-energy CT-Derived Iodine Map for Predicting PD-L1 Expression in Gastric Cancer: A Multicenter Study. Acad Radiol. 2026. PMID: 41547630. doi:10.1016/j.acra.2025.12.033

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。