Decuyper Milan, Bonte Stijn, Deblaere Karel, Van Holen Roel
Medical Image and Signal Processing (MEDISIP), Ghent University, Ghent, Belgium.
Medical Image and Signal Processing (MEDISIP), Ghent University, Ghent, Belgium.
Comput Med Imaging Graph. 2021 Mar;88:101831. doi: 10.1016/j.compmedimag.2020.101831. Epub 2020 Nov 27.
In the WHO glioma classification guidelines grade (glioblastoma versus lower-grade glioma), IDH mutation and 1p/19q co-deletion status play a central role as they are important markers for prognosis and optimal therapy planning. Currently, diagnosis requires invasive surgical procedures. Therefore, we propose an automatic segmentation and classification pipeline based on routinely acquired pre-operative MRI (T1, T1 postcontrast, T2 and/or FLAIR). A 3D U-Net was designed for segmentation and trained on the BraTS 2019 training dataset. After segmentation, the 3D tumor region of interest is extracted from the MRI and fed into a CNN to simultaneously predict grade, IDH mutation and 1p19q co-deletion. Multi-task learning allowed to handle missing labels and train one network on a large dataset of 628 patients, collected from The Cancer Imaging Archive and BraTS databases. Additionally, the network was validated on an independent dataset of 110 patients retrospectively acquired at the Ghent University Hospital (GUH). Segmentation performance calculated on the BraTS validation set shows an average whole tumor dice score of 90% and increased robustness to missing image modalities by randomly excluding input MRI during training. Classification area under the curve scores are 93%, 94% and 82% on the TCIA test data and 94%, 86% and 87% on the GUH data for grade, IDH and 1p19q status respectively. We developed a fast, automatic pipeline to segment glioma and accurately predict important (molecular) markers based on pre-therapy MRI.
在世界卫生组织的胶质瘤分类指南（胶质母细胞瘤与低级别胶质瘤）中，异柠檬酸脱氢酶（IDH）突变和1p/19q共缺失状态起着核心作用，因为它们是预后和优化治疗方案规划的重要标志物。目前，诊断需要侵入性手术操作。因此，我们提出了一种基于常规获取的术前磁共振成像（MRI）（T1加权像、T1加权增强像、T2加权像和/或液体衰减反转恢复序列像）的自动分割和分类流程。设计了一个3D U-Net用于分割，并在BraTS 2019训练数据集上进行训练。分割后，从MRI中提取3D肿瘤感兴趣区域，并将其输入到一个卷积神经网络（CNN）中，以同时预测级别、IDH突变和1p19q共缺失。多任务学习能够处理缺失标签的情况，并在从癌症成像存档库和BraTS数据库收集的628例患者的大型数据集中训练一个网络。此外，该网络在根特大学医院（GUH）回顾性获取的110例患者的独立数据集上进行了验证。在BraTS验证集上计算的分割性能显示，整个肿瘤的平均骰子系数得分达到90%，并且通过在训练期间随机排除输入MRI，提高了对缺失图像模态的鲁棒性。在TCIA测试数据上，级别、IDH和1p19q状态的曲线下面积分类得分分别为93%、94%和82%，在GUH数据上分别为94%、86%和87%。我们开发了一种快速、自动的流程来分割胶质瘤，并基于治疗前的MRI准确预测重要的（分子）标志物。
