Department of Radiology, First Affiliated Hospital With Nanjing Medical University, Nanjing, China.
Shanghai Key Laboratory of Magnetic Resonance, East China Normal University, Shanghai, China.
J Am Coll Radiol. 2019 Jul;16(7):952-960. doi: 10.1016/j.jacr.2018.12.017. Epub 2019 Feb 4.
The aim of this study was to develop and validate a computational clinical decision support system (DSS) on the basis of CT radiomics features for the prediction of lymph node (LN) metastasis in gastric cancer (GC) using machine learning-based analysis.
Clinicopathologic and CT imaging data were retrospectively collected from 490 patients who were diagnosed with GC between January 2002 and December 2016. Radiomics features were extracted from venous-phase CT images. Relevant features were selected, ranked, and modeled using a support vector machine classifier in 326 training and validation data sets. A model test was performed independently in a test set (n = 164). Finally, a head-to-head comparison of the diagnostic performance of the DSS and that of the conventional staging criterion was performed.
Two hundred ninety-seven of the 490 patients examined had histopathologic evidence of LN metastasis, yielding a 60.6% metastatic rate. The area under the curve for predicting LN+ was 0.824 (95% confidence interval, 0.804-0.847) for the DSS in the training and validation data and 0.764 (95% confidence interval, 0.699-0.833) in the test data. The calibration plots showed good concordance between the predicted and observed probability of LN+ using the DSS approach. The DSS was better able to predict LN metastasis than the conventional staging criterion in the training and validation data (accuracy 76.4% versus 63.5%) and in the test data (accuracy 71.3% versus 63.2%) CONCLUSIONS: A DSS based on 13 "worrisome" radiomics features appears to be a promising tool for the preoperative prediction of LN status in patients with GC.
本研究旨在开发和验证一种基于 CT 放射组学特征的计算临床决策支持系统（DSS），以基于机器学习的分析预测胃癌（GC）的淋巴结（LN）转移。
回顾性收集了 2002 年 1 月至 2016 年 12 月期间诊断为 GC 的 490 例患者的临床病理和 CT 影像学数据。从静脉期 CT 图像中提取放射组学特征。使用支持向量机分类器在 326 个训练和验证数据集以及 164 个测试集中选择、排序和建模相关特征。最后，对头对头比较 DSS 和传统分期标准的诊断性能。
在 490 例患者中，有 297 例经病理证实有 LN 转移，转移率为 60.6%。DSS 在训练和验证数据中的预测 LN+的 AUC 为 0.824（95%置信区间，0.804-0.847），在测试数据中的 AUC 为 0.764（95%置信区间，0.699-0.833）。校准图显示，使用 DSS 方法预测 LN+的概率与观察到的概率之间具有良好的一致性。在训练和验证数据（准确率 76.4%对 63.5%）以及测试数据（准确率 71.3%对 63.2%）中，DSS 比传统分期标准更能预测 LN 转移。
基于 13 个“令人担忧”的放射组学特征的 DSS 似乎是一种有前途的工具，可用于预测 GC 患者的 LN 状态。
