文献速递泌尿生殖腹部MR

基于多参数MRI的系统性影像组学分析术前预测膀胱癌Ki67表达与组织学分级

Systematic radiomics analysis based on multiparameter MRI to preoperatively predict the expression of Ki67 and histological grade in patients with bladder cancer.
PubMed 文献 · Br J Radiol · 2023-03-08

本研究回顾性纳入283例膀胱癌患者,基于T1WI、T2WI、DWI和DCE等多参数MRI提取瘤内及瘤周影像组学特征,采用mRMR和LASSO筛选特征并构建六种机器学习分类器。随机森林模型表现最佳,MP-MRI模型预测Ki67的AUC在训练集和测试集分别为0.977和0.852,预测组织学分级的AUC分别为0.972和0.710。影像组学有望术前预测膀胱癌多种病理结局,为临床决策提供指导。

要点

  • 瘤内特征对Ki67预测贡献更大,瘤周特征对组织学分级预测贡献更大。
  • 随机森林在预测Ki67和组织学分级中均表现最佳。
  • 不同特征筛选方法、分割区域、分类器和MRI序列均影响模型性能。

核心结论

  • 基于多参数MRI的影像组学模型可术前预测膀胱癌Ki67表达和组织学分级,随机森林模型性能最佳。

研究设计与样本

  • 回顾性研究纳入2012至2021年间283例膀胱癌患者,采用T1WI、T2WI、DWI和DCE多参数MRI序列。
  • 同时提取瘤内和瘤周区域的影像组学特征,使用mRMR和LASSO算法筛选特征。
  • 采用六种机器学习分类器构建影像组学模型,并选择最佳模型进行模型构建。

诊断标准与关键数值

  • mRMR算法更适用于Ki67预测,LASSO算法更适用于组织学分级预测。
  • MP-MRI模型预测Ki67的AUC在训练集为0.977,在测试集为0.852。
  • MP-MRI模型预测组织学分级的AUC在训练集为0.972,在测试集为0.710。

临床应用与阅片要点

  • 术前多参数MRI影像组学可无创预测膀胱癌Ki67表达和组织学分级,辅助临床决策。
  • 分析Ki67时应侧重瘤内区域特征,分析组织学分级时应关注瘤周区域特征。
  • 随机森林分类器在本研究中的预测性能最佳,可作为建模首选。

局限性

  • 本研究为回顾性设计,可能存在选择偏倚。
  • 组织学分级测试集AUC为0.710,泛化性能有待提升。

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: Bladder cancer is among the most prevalent urothelial malignancies. Radiomics-based preoperative prediction of Ki67 and histological grade will facilitate clinical decision-making.

METHODS: This retrospective study recruited 283 bladder cancer patients between 2012 and 2021. Multiparameter MRI sequences included: T1WI, T2WI, diffusion-weighted imaging (DWI), and dynamic contrast-enhanced (DCE) imaging. The radiomics features of intratumoral and peritumoral regions were extracted simultaneously. Max-Relevance and Min-Redundancy (mRMR) and least absolute shrinkage and selection operator (LASSO) algorithms were employed to select the features. Six machine learning-based classifiers were adopted to construct the radiomics models, and the best was chosen for the model construction.

RESULTS: The mRMR and LASSO algorithms were more suitable for Ki67 and histological grade, respectively. Additionally, Ki67 had a higher proportion of intratumoral features, while peritumoral features accounted for a greater proportion of the histological grade. Random forests performed the best in predicting both pathological outcomes. Consequently, the multiparameter MRI (MP-MRI) models achieved area under the curve (AUC) values of 0.977 and 0.852 for Ki67 in training and test sets, respectively, and 0.972 and 0.710 for the histological grade.

CONCLUSION: Radiomics holds the potential to predict multiple pathological outcomes of bladder cancer preoperatively and are expected to provide clinical decision-making guidance. Furthermore, our work inspired the process of radiomics research.

ADVANCES IN KNOWLEDGE: This study demonstrated that different feature selection techniques, segmentation regions, classifiers, and MRI sequences will affect the performance of the model. We systematically demonstrated that radiomics can predict histological grade and Ki67.

来源:Br J Radiol,查看原文。版权归原作者所有。

引用

  1. Fan, Yu, Ni, et al. Systematic radiomics analysis based on multiparameter MRI to preoperatively predict the expression of Ki67 and histological grade in patients with bladder cancer. Br J Radiol. 2023. PMID: 36883677. doi:10.1259/bjr.20221086

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。