文献速递腹部技术与安全CT

级联分割-分类深度学习框架用于晚期胃癌术前预测隐匿性腹膜转移和早期复发

A Cascaded Segmentation-Classification Deep Learning Framework for Preoperative Prediction of Occult Peritoneal Metastasis and Early Recurrence in Advanced Gastric Cancer.
PubMed 文献 · Acad Radiol · 2025-09-04

该多中心研究纳入3家机构765例晚期胃癌患者,构建基于CT的级联深度学习框架(V-Net肿瘤分割+深度学习转移风险分类),结合临床病理因素建立联合模型。联合模型预测隐匿性腹膜转移的AUC为0.938(训练集)和0.916(内部验证集),预测早期腹膜复发的AUC为0.820-0.825,敏感度79.7-88.9%、特异度72.4-73.3%。

要点

  • 联合模型预测OPM的AUC在训练集为0.938、内部验证集为0.916,优于临床模型和纯深度学习模型。
  • 联合模型预测早期腹膜复发的AUC为0.820-0.825,敏感度79.7-88.9%、特异度72.4-73.3%。
  • Borrmann分型、CA125≥35U/mL和CT-N+分期是OPM的独立预测因素。

核心结论

  • 基于CT的级联框架可对晚期胃癌的隐匿性腹膜转移和早期腹膜复发进行可靠的术前风险分层,可能优化个体化治疗路径的适应证。

研究设计与样本

  • 多中心研究,纳入3家机构765例晚期胃癌患者,构建两阶段框架:基于CT的V-Net肿瘤分割和基于分割肿瘤区域的深度学习转移风险分类,并将临床病理预测因素与深度学习概率整合构建联合模型;验证队列包括内部验证(Test1用于OPM,n=168;Test2用于早期PR,n=212)和外部验证(Test3用于早期PR,n=57,来自两个独立中心)。

诊断标准与关键数值

  • 联合模型预测OPM的AUC:训练集0.938,Test1 0.916;较临床模型提高∆AUC +0.039至+0.107,较纯深度学习模型提高∆AUC +0.044至+0.104。
  • 联合模型预测早期PR的AUC:Test2和Test3为0.820-0.825;敏感度79.7-88.9%,特异度72.4-73.3%。
  • 多变量分析显示Borrmann分型(OR=1.314,95%CI 1.239-1.394)、CA125≥35U/mL(OR=1.301,95%CI 1.127-1.499)和CT-N+分期(OR=1.259,95%CI 1.124-1.415)为OPM独立预测因素。

临床应用与阅片要点

  • 联合模型可用于晚期胃癌术前OPM和早期腹膜复发的风险分层,决策曲线分析证实其在临床阈值范围内有净临床获益。
  • 作者认为该框架可能优化个体化治疗路径的适应证。

局限性

  • 原文未报告局限性。

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

PURPOSE: To develop a cascaded deep learning (DL) framework integrating tumor segmentation with metastatic risk stratification for preoperative prediction of occult peritoneal metastasis (OPM) in advanced gastric cancer (GC), and validate its generalizability for early peritoneal recurrence (PR) prediction.

METHODS: This multicenter study enrolled 765 patients with advanced GC from three institutions. We developed a two-stage framework as follows: (1) V-Net-based tumor segmentation on CT; (2) DL-based metastatic risk classification using segmented tumor regions. Clinicopathological predictors were integrated with deep learning probabilities to construct a combined model. Validation cohorts comprised: Internal validation (Test1 for OPM, n=168; Test2 for early PR, n=212) and External validation (Test3 for early PR, n=57 from two independent centers).

RESULTS: Multivariable analysis identified Borrmann type (OR=1.314, 95% CI: 1.239-1.394), CA125 ≥35U/mL (OR=1.301, 95% CI: 1.127-1.499), and CT-N+ stage (OR=1.259, 95% CI: 1.124-1.415) as independent OPM predictors. The combined model demonstrated robust performance for both OPM and early PR prediction: achieving AUCs of 0.938 (Train) and 0.916 (Test1) for OPM with improvements over clinical (∆AUC +0.039-+0.107) and DL-only models (∆AUC +0.044-+0.104), while attaining AUC 0.820-0.825 for early PR (Test2 and Test3) with balanced sensitivity (79.7-88.9%) and specificity (72.4-73.3%). Decision curve analysis confirmed net clinical benefit across clinical thresholds.

CONCLUSION: This CT-based cascaded framework enables reliable preoperative risk stratification for OPM and early PR in advanced GC, potentially refining indications for personalized therapeutic pathways.

来源:Acad Radiol,查看原文。版权归原作者所有。

引用

  1. Zou T, Chen P, Wang T, et al. A Cascaded Segmentation-Classification Deep Learning Framework for Preoperative Prediction of Occult Peritoneal Metastasis and Early Recurrence in Advanced Gastric Cancer. Acad Radiol. 2025. PMID: 40912952. doi:10.1016/j.acra.2025.08.006

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。