文献速递神经MR

知识引导的无钆MRI影像组学预测IDH突变型成人弥漫性胶质瘤1p/19q共缺失:一项双中心研究

Knowledge-guided gadolinium-free MRI radiomics predict 1p/19q co-deletion in IDH-mutant adult-type diffuse gliomas: A dual-center study.
PubMed 文献 · Eur J Radiol · 2025-11-22

本研究将视觉特征知识融入影像组学,基于无钆增强MRI构建可解释模型,用于术前预测IDH突变型成人弥漫性胶质瘤的1p/19q共缺失。知识引导的支持向量机模型在内外部验证集AUC分别为0.893和0.839,优于常规影像组学模型(外部集AUC 0.681-0.767),在多数性能指标上差异有统计学意义(p<0.01)。

要点

  • 知识引导的支持向量机模型外部验证AUC为0.839,优于常规影像组学模型(外部集AUC 0.681-0.767)。
  • 模型基于无钆增强的T1WI、T2WI和FLAIR序列,结合T2-FLAIR不匹配、均匀性、强化和钙化四个视觉特征。
  • 知识引导模型在多数性能指标上优于对应常规影像组学模型(p<0.01),跨中心泛化性更好。

核心结论

  • 知识引导的影像组学框架提高了IDH突变型成人弥漫性胶质瘤1p/19q共缺失的术前预测能力,并增强了模型可解释性和跨中心泛化性。

研究设计与样本

  • 回顾性双中心研究,纳入215例经手术证实的IDH突变型成人弥漫性胶质瘤患者,按训练集、内部验证集和外部验证集划分;基于T1WI、T2WI和FLAIR提取高读者间稳定性特征,采用层次聚类和最大相关最小冗余消除共线性和冗余,使用Lasso进行稳定性特征选择,训练支持向量机和极端梯度提升两种分类器,并采用嵌套交叉验证;以视觉特征(T2-FLAIR不匹配、均匀性、强化、钙化)为知识引导,与无视觉引导的常规影像组学模型比较。

诊断标准与关键数值

  • 知识引导支持向量机模型内部验证集AUC为0.893(95%CI:0.859-0.928)。
  • 知识引导支持向量机模型外部验证集AUC为0.839(95%CI:0.794-0.885)。
  • 常规影像组学模型外部验证集AUC为0.681-0.767。
  • 知识引导模型在验证集多数性能指标上优于对应常规影像组学模型(p<0.01)。

临床应用与阅片要点

  • 作者认为知识引导影像组学框架可提高术前预测IDH突变型成人弥漫性胶质瘤1p/19q共缺失的可解释性和跨中心泛化性。

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: By integrating visual knowledge into radiomics, we aim to develop an interpretable and robust radiomic model based on Gadolinium-free MRI for preoperative identification of 1p/19q co-deletion in IDH-mutant adult-type diffuse gliomas.

MATERIALS AND METHODS: MRI from 215 surgically confirmed IDH-mutant Adult-type Diffuse Glioma patients were collected from two centers and split into training, internal and external validation set. For knowledge-guided radiomics, features driven from T1WI, T2WI, and FLAIR demonstrating high interreader stability were extracted from the training set. Features significantly associated with four visual features (T2-FLAIR mismatch, homogeneity, enhancement, and calcification) were identified, and hierarchical clustering and Max-Relevance and Min-Redundancy were used to eliminate multicollinearity and redundancy. Stability selection using Lasso was performed for robust feature selection. Two classifiers (Support Vector Machines and Extreme Gradient Boosting) were trained with nested cross-validation for 1p/19q co-deletion prediction. For comparison, regular radiomic models without visual guidance were developed to evaluate the superiority of the knowledge-guided radiomics. This retrospective study was approved by the institutional review board of the centers and the requirement for informed consent was waived.

RESULTS: The knowledge-guided Support Vector Machines model achieved AUCs of 0.893 (95 %CI: 0.859-0.928) and 0.839 (95 % CI: 0.794-0.885) in the internal and external validation sets. The two regular radiomics models demonstrated various performance degradation, especially in the external set (AUCs: 0.681-0.767). The knowledge-guided radiomics model of either classifier outperformed corresponding regular radiomics across most performance metrics (p < 0.01) in the validation sets.

CONCLUSIONS: The knowledge-guided radiomic framework improves preoperative prediction of 1p/19q co-deletion in IDH-mutant adult-type diffuse gliomas, offering enhanced model interpretability and cross-center generalizability.

来源:Eur J Radiol,查看原文。版权归原作者所有。

引用

  1. He W, Chen L, Leng X, et al. Knowledge-guided gadolinium-free MRI radiomics predict 1p/19q co-deletion in IDH-mutant adult-type diffuse gliomas: A dual-center study. Eur J Radiol. 2025. PMID: 41308577. doi:10.1016/j.ejrad.2025.112569

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。