Buda Mateusz, AlBadawy Ehab A, Saha Ashirbani, Mazurowski Maciej A
Department of Radiology, Duke University School of Medicine, 2424 Erwin Rd, Suite 302, Durham, NC 27705 (M.B., E.A.A., A.S., M.A.M.); Department of Electrical and Computer Engineering, Duke University, Durham, NC (M.A.M.); and Department of Biostatistics and Bioinformatics, Duke University, Durham, NC (M.A.M.).
Radiol Artif Intell. 2020 Jan 29;2(1):e180050. doi: 10.1148/ryai.2019180050. eCollection 2020 Jan.
To employ deep learning to predict genomic subtypes of lower-grade glioma (LLG) tumors based on their appearance at MRI.
Imaging data from The Cancer Imaging Archive and genomic data from The Cancer Genome Atlas from 110 patients from five institutions with lower-grade gliomas (World Health Organization grade II and III) were used in this study. A convolutional neural network was trained to predict tumor genomic subtype based on the MRI of the tumor. Two different deep learning approaches were tested: training from random initialization and transfer learning. Deep learning models were pretrained on glioblastoma MRI, instead of natural images, to determine if performance was improved for the detection of LGGs. The models were evaluated using area under the receiver operating characteristic curve (AUC) with cross-validation. Imaging data and annotations used in this study are publicly available.
The best performing model was based on transfer learning from glioblastoma MRI. It achieved AUC of 0.730 (95% confidence interval [CI]: 0.605, 0.844) for discriminating cluster-of-clusters 2 from others. For the same task, a network trained from scratch achieved an AUC of 0.680 (95% CI: 0.538, 0.811), whereas a model pretrained on natural images achieved an AUC of 0.640 (95% CI: 0.521, 0.763).
These findings show the potential of utilizing deep learning to identify relationships between cancer imaging and cancer genomics in LGGs. However, more accurate models are needed to justify clinical use of such tools, which might be obtained using substantially larger training datasets.© RSNA, 2020.
运用深度学习基于低级别胶质瘤（LLG）肿瘤的磁共振成像（MRI）表现预测其基因组亚型。
本研究使用了来自五个机构的110例低级别胶质瘤（世界卫生组织二级和三级）患者的癌症影像存档（The Cancer Imaging Archive）的影像数据以及癌症基因组图谱（The Cancer Genome Atlas）的基因组数据。训练了一个卷积神经网络，基于肿瘤的MRI预测肿瘤基因组亚型。测试了两种不同的深度学习方法：随机初始化训练和迁移学习。深度学习模型在胶质母细胞瘤MRI而非自然图像上进行预训练，以确定对低级别胶质瘤检测的性能是否有所提高。使用受试者操作特征曲线下面积（AUC）和交叉验证对模型进行评估。本研究中使用的影像数据和注释均可公开获取。
表现最佳的模型基于从胶质母细胞瘤MRI的迁移学习。在区分簇状簇2与其他类型时，其AUC为0.730（95%置信区间[CI]：[0.605, 0.844]）。对于相同任务，从零开始训练的网络AUC为0.680（95% CI：[0.538, 0.811]），而在自然图像上预训练的模型AUC为0.640（95% CI：[0.521, 0.763]）。
这些发现表明利用深度学习识别低级别胶质瘤中癌症影像与癌症基因组学之间关系的潜力。然而，需要更精确的模型来证明此类工具的临床应用合理性，这可能需要使用实质上更大的训练数据集来实现。© RSNA，2020年。
