文献速递胸部CT

基于机器学习的CT影像组学回归模型用于预测肺功能

Machine-Learning-Based Computed Tomography Radiomics Regression Model for Predicting Pulmonary Function.
PubMed 文献 · Acad Radiol · 2025-04-10

本回顾性研究纳入2018年1月至2024年4月间接受胸部CT和肺功能检查的患者1585例(含外部病例102例),构建并验证机器学习回归模型预测用力肺活量和第1秒用力呼气容积,联合全肺影像组学与临床特征的模型表现最佳,外部测试集预测FVC的CCC为0.745、R²为0.601,预测FEV1的CCC为0.744、R²为0.527。

要点

  • 联合全肺影像组学与临床特征的模型在训练、验证、测试及外部测试集中预测FVC与FEV1均表现最佳。
  • 年龄、性别和肺气肿是预测FVC与FEV1的重要影响因素,各自还有不同的影像组学特征参与贡献。
  • 外部测试集预测FVC的CCC为0.745(95%CI 0.642-0.818)、R²为0.601(95%CI 0.453-0.707)。

核心结论

  • 全肺影像组学特征可用于构建回归模型以改善肺功能预测。

研究设计与样本

  • 回顾性研究,纳入2018年1月至2024年4月间接受胸部CT和肺功能检查的患者1585例,其中外部病例102例,构建并验证机器学习回归模型预测FVC和FEV1,模型纳入全肺影像组学与临床特征,以肺量计检查结果为参照。

诊断标准与关键数值

  • 联合模型在外部测试集预测FVC:CCC 0.745(95%CI 0.642-0.818),R² 0.601(95%CI 0.453-0.707)。
  • 联合模型在外部测试集预测FEV1:CCC 0.744(95%CI 0.633-0.824),R² 0.527(95%CI 0.298-0.675)。
  • 模型性能采用平均绝对误差、均方误差、均方根误差、一致性相关系数(CCC)和R²值评估,并与肺量计检查结果比较。

临床应用与阅片要点

  • 年龄、性别和肺气肿是预测FVC与FEV1的重要影响因素,阅片时需关注肺气肿征象并纳入模型解读。
  • 全肺影像组学特征与临床特征联合建模可改善肺功能预测,模型决策可通过基于SHapley Additive exPlanations的可解释方法进行个体化分析。

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

RATIONALE AND OBJECTIVES: Chest computed tomography (CT) radiomics can be utilized for categorical predictions; however, models predicting pulmonary function indices directly are lacking. This study aimed to develop machine-learning-based regression models to predict pulmonary function using chest CT radiomics.

METHODS: This retrospective study enrolled patients who underwent chest CT and pulmonary function tests between January 2018 and April 2024. Machine-learning regression models were constructed and validated to predict pulmonary function indices, including forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1). The models incorporated radiomics of the whole lung and clinical features. Model performance was evaluated using mean absolute error, mean squared error, root mean squared error, concordance correlation coefficient (CCC), and R-squared (R2) value and compared to spirometry results. Individual explanations of the models' decisions were analyzed using an explainable approach based on SHapley Additive exPlanations.

RESULTS: In total, 1585 cases were included in the analysis, with 102 of them being external cases. Across the training, validation, test, and external test sets, the combined model consistently achieved the best performance in the regression task for predicting FVC (e.g. external test set: CCC, 0.745 [95% confidence interval 0.642-0.818]; R2, 0.601 [0.453-0.707]) and FEV1 (e.g. external test set: CCC, 0.744 [0.633-0.824]; R2, 0.527 [0.298-0.675]). Age, sex, and emphysema were important factors for both FVC and FEV1, while distinct radiomics features contributed to each.

CONCLUSION: Whole-lung-based radiomics features can be used to construct regression models to improve pulmonary function prediction.

来源:Acad Radiol,查看原文。版权归原作者所有。

引用

  1. Wang W, Sun Y, Wu R, et al. Machine-Learning-Based Computed Tomography Radiomics Regression Model for Predicting Pulmonary Function. Acad Radiol. 2025. PMID: 40216578. doi:10.1016/j.acra.2025.03.038

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。