Department of Radiology, Severance Hospital, Research Institute of Radiological Science and Center for Clinical Image Data Science, Yonsei University College of Medicine, Seoul, South Korea.
Department of Radiology and Research Institute, National Health Insurance Service Ilsan Hospital, Goyang, South Korea.
J Neurooncol. 2021 Dec;155(3):267-276. doi: 10.1007/s11060-021-03870-z. Epub 2021 Oct 14.
In glioma, molecular alterations are closely associated with disease prognosis. This study aimed to develop a radiomics-based multiple gene prediction model incorporating mutual information of each genetic alteration in glioblastoma and grade 4 astrocytoma, IDH-mutant.
From December 2014 through January 2020, we enrolled 418 patients with pathologically confirmed glioblastoma (based on the 2016 WHO classification). All selected patients had preoperative MRI and isocitrate dehydrogenase (IDH) mutation, O-6-methylguanine-DNA methyltransferase (MGMT) promoter methylation, epidermal growth factor receptor amplification, and alpha-thalassemia/mental retardation syndrome X-linked (ATRX) loss status. Patients were randomly split into training and test sets (7:3 ratio). Enhancing tumor and peritumoral T2-hyperintensity were auto-segmented, and 660 radiomics features were extracted. We built binary relevance (BR) and ensemble classifier chain (ECC) models for multi-label classification and compared their performance. In the classifier chain, we calculated the mean absolute Shapley value of input features.
The micro-averaged area under the curves (AUCs) for the test set were 0.804 and 0.842 in BR and ECC models, respectively. IDH mutation status was predicted with the highest AUCs of 0.964 (BR) and 0.967 (ECC). The ECC model showed higher AUCs than the BR model for ATRX (0.822 vs. 0.775) and MGMT promoter methylation (0.761 vs. 0.653) predictions. The mean absolute Shapley values suggested that predicted outcomes from the prior classifiers were important for better subsequent predictions along the classifier chains.
We built a radiomics-based multiple gene prediction chained model that incorporates mutual information of each genetic alteration in glioblastoma and grade 4 astrocytoma, IDH-mutant and performs better than a simple bundle of binary classifiers using prior classifiers' prediction probability.
在神经胶质瘤中，分子改变与疾病预后密切相关。本研究旨在开发一种基于放射组学的多基因预测模型，该模型纳入了 IDH 突变型胶质母细胞瘤和 4 级星形细胞瘤中每种遗传改变的互信息。
本研究纳入了 2016 年 WHO 分类标准确诊的 418 例胶质母细胞瘤患者，所有患者术前均行 MRI 和异柠檬酸脱氢酶（IDH）突变、O-6-甲基鸟嘌呤-DNA 甲基转移酶（MGMT）启动子甲基化、表皮生长因子受体扩增以及 X 连锁α-地中海贫血/智力低下综合征（ATRX）缺失状态检测。患者被随机分为训练集和测试集（7:3 比例）。自动分割增强肿瘤和瘤周 T2 高信号区，并提取 660 个放射组学特征。我们构建了二元相关性（BR）和集成分类器链（ECC）模型进行多标签分类，并比较了它们的性能。在分类器链中，我们计算了输入特征的平均绝对 Shapley 值。
测试集的微平均曲线下面积（AUC）分别为 BR 模型 0.804 和 ECC 模型 0.842。IDH 突变状态的预测 AUC 最高，分别为 BR 模型 0.964 和 ECC 模型 0.967。ECC 模型在 ATRX（0.822 vs. 0.775）和 MGMT 启动子甲基化（0.761 vs. 0.653）预测中的 AUC 均高于 BR 模型。平均绝对 Shapley 值表明，来自前序分类器的预测结果对沿着分类器链的后续更好预测很重要。
我们构建了一种基于放射组学的多基因预测链式模型，该模型纳入了 IDH 突变型胶质母细胞瘤和 4 级星形细胞瘤中每种遗传改变的互信息，且表现优于使用前序分类器预测概率的简单二元分类器捆绑。
