Support Center for Advanced Neuroimaging, University Institute for Diagnostic and Interventional Neuroradiology, Inselspital, Bern University Hospital, Bern, Switzerland.
Swiss Paraplegic Centre, Nottwil, Switzerland.
Sci Rep. 2021 Jan 13;11(1):1087. doi: 10.1038/s41598-020-79925-4.
Segmentation of white matter lesions and deep grey matter structures is an important task in the quantification of magnetic resonance imaging in multiple sclerosis. In this paper we explore segmentation solutions based on convolutional neural networks (CNNs) for providing fast, reliable segmentations of lesions and grey-matter structures in multi-modal MR imaging, and the performance of these methods when applied to out-of-centre data. We trained two state-of-the-art fully convolutional CNN architectures on the 2016 MSSEG training dataset, which was annotated by seven independent human raters: a reference implementation of a 3D Unet, and a more recently proposed 3D-to-2D architecture (DeepSCAN). We then retrained those methods on a larger dataset from a single centre, with and without labels for other brain structures. We quantified changes in performance owing to dataset shift, and changes in performance by adding the additional brain-structure labels. We also compared performance with freely available reference methods. Both fully-convolutional CNN methods substantially outperform other approaches in the literature when trained and evaluated in cross-validation on the MSSEG dataset, showing agreement with human raters in the range of human inter-rater variability. Both architectures showed drops in performance when trained on single-centre data and tested on the MSSEG dataset. When trained with the addition of weak anatomical labels derived from Freesurfer, the performance of the 3D Unet degraded, while the performance of the DeepSCAN net improved. Overall, the DeepSCAN network predicting both lesion and anatomical labels was the best-performing network examined.
对磁共振成像中的白质病变和深部灰质结构进行分割是多发性硬化症量化分析的一项重要任务。在本文中，我们探索了基于卷积神经网络（CNN）的分割解决方案，以便为多模态磁共振成像中的病变和灰质结构提供快速、可靠的分割，并研究这些方法应用于中心外数据时的性能。我们在 2016 年 MSSEG 训练数据集上对两个最先进的全卷积 CNN 架构进行了训练，该数据集由七位独立的人类评估者进行了注释：3D Unet 的参考实现和最近提出的 3D-to-2D 架构（DeepSCAN）。然后，我们在一个来自单一中心的更大数据集上重新训练了这些方法，其中包括或不包括其他脑结构的标签。我们量化了由于数据集偏移和添加其他脑结构标签而导致的性能变化。我们还将性能与免费提供的参考方法进行了比较。在 MSSEG 数据集上进行交叉验证时，这两种全卷积 CNN 方法在训练和评估时都明显优于文献中的其他方法，与人类评估者的一致性在人类评估者之间的可变性范围内。当在单中心数据上进行训练并在 MSSEG 数据集上进行测试时，两种架构的性能都有所下降。当使用来自 Freesurfer 的弱解剖标签进行训练时，3D Unet 的性能下降，而 DeepSCAN 网络的性能提高。总体而言，预测病变和解剖标签的 DeepSCAN 网络是表现最佳的网络。
