Xu Zhenlin, Niethammer Marc
University of North Carolina, Chapel Hill, NC, USA.
Med Image Comput Comput Assist Interv. 2019 Oct;11765:420-429. doi: 10.1007/978-3-030-32245-8_47. Epub 2019 Oct 10.
Deep convolutional neural networks (CNNs) are state-of-the-art for semantic image segmentation, but typically require many labeled training samples. Obtaining 3D segmentations of medical images for supervised training is difficult and labor intensive. Motivated by classical approaches for joint segmentation and registration we therefore propose a deep learning framework that jointly learns networks for image registration and image segmentation. In contrast to previous work on deep unsupervised image registration, which showed the benefit of weak supervision via image segmentations, our approach can use existing segmentations when available and computes them via the segmentation network otherwise, thereby providing the same registration benefit. Conversely, segmentation network training benefits from the registration, which essentially provides a realistic form of data augmentation. Experiments on knee and brain 3D magnetic resonance (MR) images show that our approach achieves large simultaneous improvements of segmentation and registration accuracy (over independently trained networks) and allows training high-quality models with very limited training data. Specifically, in a one-shot-scenario (with only one manually labeled image) our approach increases Dice scores (%) over an unsupervised registration network by 2.7 and 1.8 on the knee and brain images respectively.
深度卷积神经网络（CNN）在语义图像分割方面处于领先水平，但通常需要大量带标签的训练样本。获取用于监督训练的医学图像的3D分割很困难且耗费人力。因此，受联合分割和配准的经典方法启发，我们提出了一个深度学习框架，该框架联合学习用于图像配准和图像分割的网络。与先前关于深度无监督图像配准的工作不同，先前的工作通过图像分割展示了弱监督的好处，我们的方法在有可用的现有分割时可以使用它们，否则通过分割网络计算分割，从而提供相同的配准优势。相反，分割网络的训练受益于配准，配准本质上提供了一种现实的数据增强形式。在膝盖和大脑的3D磁共振（MR）图像上的实验表明，我们的方法在分割和配准精度方面实现了大幅同步提升（相对于独立训练的网络），并且能够使用非常有限的训练数据训练高质量模型。具体而言，在单样本场景（仅一张手动标注的图像）中，我们的方法在膝盖和大脑图像上分别比无监督配准网络的骰子系数得分（%）提高了2.7和1.8。
