Institute of Computer Graphics and Vision, Graz University of Technology, Graz, Austria.
Ludwig Boltzmann Institute for Clinical Forensic Imaging, Graz, Austria.
Med Image Anal. 2019 May;54:207-219. doi: 10.1016/j.media.2019.03.007. Epub 2019 Mar 25.
In many medical image analysis applications, only a limited amount of training data is available due to the costs of image acquisition and the large manual annotation effort required from experts. Training recent state-of-the-art machine learning methods like convolutional neural networks (CNNs) from small datasets is a challenging task. In this work on anatomical landmark localization, we propose a CNN architecture that learns to split the localization task into two simpler sub-problems, reducing the overall need for large training datasets. Our fully convolutional SpatialConfiguration-Net (SCN) learns this simplification due to multiplying the heatmap predictions of its two components and by training the network in an end-to-end manner. Thus, the SCN dedicates one component to locally accurate but ambiguous candidate predictions, while the other component improves robustness to ambiguities by incorporating the spatial configuration of landmarks. In our extensive experimental evaluation, we show that the proposed SCN outperforms related methods in terms of landmark localization error on a variety of size-limited 2D and 3D landmark localization datasets, i.e., hand radiographs, lateral cephalograms, hand MRIs, and spine CTs.
在许多医学图像分析应用中，由于图像采集的成本和专家所需的大量手动注释工作，只有有限数量的训练数据可用。从小数据集训练最新的深度学习方法（如卷积神经网络（CNN））是一项具有挑战性的任务。在这项关于解剖学标志定位的工作中，我们提出了一种 CNN 架构，该架构可将定位任务分为两个更简单的子问题，从而减少对大型训练数据集的总体需求。我们的完全卷积空间配置网络（SCN）通过将其两个组件的热图预测相乘并以端到端的方式训练网络，从而学会了这种简化。因此，SCN 专门为局部准确但存在歧义的候选预测分配一个组件，而另一个组件通过合并地标空间配置来提高对歧义的鲁棒性。在我们广泛的实验评估中，我们表明，所提出的 SCN 在各种大小受限的 2D 和 3D 地标定位数据集（即手部 X 光片、侧位头颅 X 光片、手部 MRI 和脊柱 CT）上的地标定位误差方面优于相关方法。
