用于预测异柠檬酸脱氢酶野生型、端粒酶逆转录酶启动子突变型胶质瘤的磁共振成像变压器深度学习与放射组学

MRI transformer deep learning and radiomics for predicting IDH wild type TERT promoter mutant gliomas.

作者信息

Niu Wenju, Yan Junyu, Hao Min, Zhang Yibo, Li Tianshi, Liu Chen, Li Qijian, Liu Zihao, Su Yincheng, Peng Bo, Tan Yan, Wang Xiaochun, Wang Lei, Zhang Hui, Yang Guoqiang

机构信息

Department of Radiology, First Hospital of Shanxi Medical University, Taiyuan, 030001, China.

College of Medical Imaging, Shanxi Medical University, Taiyuan, 030001, China.

出版信息

NPJ Precis Oncol. 2025 Mar 27;9(1):89. doi: 10.1038/s41698-025-00884-y.

This study aims to predict IDH wt with TERTp-mut gliomas using multiparametric MRI sequences through a novel fusion model, while matching model classification metrics with patient risk stratification aids in crafting personalized diagnostic and prognosis evaluations.Preoperative T1CE and T2FLAIR sequences from 1185 glioma patients were analyzed. A MultiChannel_2.5D_DL model and a 2D DL model, both based on the cross-scale attention vision transformer (CrossFormer) neural network, along with a Radiomics model, were developed. These were integrated via ensemble learning into a stacking model. The MultiChannel_2.5D_DL model outperformed the 2D_DL and Radiomics models, with AUCs of 0.806-0.870. The stacking model achieved the highest AUC (0.855-0.904) across validation sets. Patients were stratified into high-risk and low-risk groups based on stacking model scores, with significant survival differences observed via Kaplan-Meier analysis and log-rank tests. The stacking model effectively identifies IDH wt TERTp-mutant gliomas and stratifies patient risk, aiding personalized prognosis.

本研究旨在通过一种新型融合模型，利用多参数MRI序列预测异柠檬酸脱氢酶（IDH）野生型伴端粒酶逆转录酶（TERTp）突变的胶质瘤，同时将模型分类指标与患者风险分层相匹配，有助于制定个性化的诊断和预后评估。分析了1185例胶质瘤患者的术前T1增强（T1CE）和液体衰减反转恢复（T2FLAIR）序列。开发了基于跨尺度注意力视觉Transformer（CrossFormer）神经网络的多通道2.5D深度学习（MultiChannel_2.5D_DL）模型和2D深度学习（2D_DL）模型，以及一个放射组学模型。通过集成学习将这些模型整合到一个堆叠模型中。MultiChannel_2.5D_DL模型的表现优于2D_DL模型和放射组学模型，曲线下面积（AUC）为0.806 - 0.870。堆叠模型在验证集中获得了最高的AUC（0.855 - 0.904）。根据堆叠模型得分将患者分为高风险和低风险组，通过Kaplan-Meier分析和对数秩检验观察到显著的生存差异。堆叠模型有效地识别出IDH野生型TERTp突变型胶质瘤并对患者风险进行分层，有助于个性化预后评估。