胶质瘤的准确分类对于免疫治疗的选择至关重要，MRI 包含大量放射组学特征，可能会提示一些预后相关信号。我们的目标是利用放射组学特征预测神经胶质瘤的新亚型，并表征其生存、免疫、基因组谱和药物反应。我们最初从 CPTAC 数据集中获得了 36 名患者的 341 张图像，用于开发深度学习模型。进一步收集了来自 TCGA_GBM 的 111 名患者的 1812 张图像和来自 TCGA_LGG 的 53 名患者的 152 张图像进行测试和验证。开发了基于 Mask R-CNN 的深度学习方法来识别神经胶质瘤患者的新亚型，并比较不同亚型的生存状态、免疫浸润模式、基因组特征、特定药物和预测模型。 200 名神经胶质瘤患者（平均年龄 33 岁）年 ± 19 [标准差]) 被登记。深度学习模型识别肿瘤区域的准确性在测试集中达到 88.3% (98/111)，在验证集中达到 83% (44/53)。根据显示不同预后结果的放射组学特征，将样本分为两个亚型（风险比，2.70）。根据免疫浸润分析结果，预后较差的亚型被定义为免疫沉默放射组学（ISR）亚型（n = 43），另一亚型为免疫激活放射组学（IAR）亚型（n = 53）。亚型特异性基因组特征将细胞系区分为 ISR 细胞系 (n = 9) 和对照细胞系 (n = 13)，并鉴定了八种 ISR 特异性药物，其中四种经过 OCTAD 数据库验证。三种基于机器学习的分类器表明，放射组学和基因组学的共同特征可以更好地预测神经胶质瘤的放射组学亚型。这些发现为放射基因组学如何识别以非侵入性方式预测预后、免疫和药物敏感性的特定亚型提供了见解。版权所有 © 2024 Elsevier Ltd. 保留所有权利。

Accurate classification of gliomas is critical to the selection of immunotherapy, and MRI contains a large number of radiomic features that may suggest some prognostic relevant signals. We aim to predict new subtypes of gliomas using radiomic features and characterize their survival, immune, genomic profiles and drug response.We initially obtained 341 images of 36 patients from the CPTAC dataset for the development of deep learning models. Further 1812 images of 111 patients from TCGA_GBM and 152 images of 53 patients from TCGA_LGG were collected for testing and validation. A deep learning method based on Mask R-CNN was developed to identify new subtypes of glioma patients and compared the survival status, immune infiltration patterns, genomic signatures, specific drugs, and predictive models of different subtypes.200 glioma patients (mean age, 33 years ± 19 [standard deviation]) were enrolled. The accuracy of the deep learning model for identifying tumor regions achieved 88.3 % (98/111) in the test set and 83 % (44/53) in the validation set. The sample was divided into two subtypes based on radiomic features showed different prognostic outcomes (hazard ratio, 2.70). According to the results of the immune infiltration analysis, the subtype with a poorer prognosis was defined as the immunosilencing radiomic (ISR) subtype (n = 43), and the other subtype was the immunoactivated radiomic (IAR) subtype (n = 53). Subtype-specific genomic signatures distinguished celllines into ISR celllines (n = 9) and control celllines (n = 13), and identified eight ISR-specific drugs, four of which were validated by the OCTAD database. Three machine learning-based classifiers showed that radiomic and genomic co-features better predicted the radiomic subtypes of gliomas.These findings provide insights into how radiogenomic could identify specific subtypes that predict prognosis, immune and drug sensitivity in a non-invasive manner.Copyright © 2024 Elsevier Ltd. All rights reserved.