Liu Xiaofeng, Xing Fangxu, Shusharina Nadya, Lim Ruth, Kuo C-C Jay, El Fakhri Georges, Woo Jonghye
Gordon Center for Medical Imaging, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, 02114.
Division of Radiation Biophysics, Department of radiation Oncology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, 02114.
Med Image Comput Comput Assist Interv. 2022 Sep;13435:66-76. doi: 10.1007/978-3-031-16443-9_7. Epub 2022 Sep 16.
Unsupervised domain adaptation (UDA) has been vastly explored to alleviate domain shifts between source and target domains, by applying a well-performed model in an unlabeled target domain via supervision of a labeled source domain. Recent literature, however, has indicated that the performance is still far from satisfactory in the presence of significant domain shifts. Nonetheless, delineating a few target samples is usually manageable and particularly worthwhile, due to the substantial performance gain. Inspired by this, we aim to develop semi-supervised domain adaptation (SSDA) for medical image segmentation, which is largely underexplored. We, thus, propose to exploit both labeled source and target domain data, in addition to unlabeled target data in a unified manner. Specifically, we present a novel asymmetric co-training (ACT) framework to integrate these subsets and avoid the domination of the source domain data. Following a divide-and-conquer strategy, we explicitly decouple the label supervisions in SSDA into two asymmetric sub-tasks, including semi-supervised learning (SSL) and UDA, and leverage different knowledge from two segmentors to take into account the distinction between the source and target label supervisions. The knowledge learned in the two modules is then adaptively integrated with ACT, by iteratively teaching each other, based on the confidence-aware pseudo-label. In addition, pseudo label noise is well-controlled with an exponential MixUp decay scheme for smooth propagation. Experiments on cross-modality brain tumor MRI segmentation tasks using the BraTS18 database showed, even with limited labeled target samples, ACT yielded marked improvements over UDA and state-of-the-art SSDA methods and approached an "upper bound" of supervised joint training.
无监督域适应（UDA）已被广泛研究，旨在通过在有标签的源域监督下，将一个性能良好的模型应用于无标签的目标域，来缓解源域和目标域之间的域偏移。然而，最近的文献表明，在存在显著域偏移的情况下，其性能仍远不能令人满意。尽管如此，由于能带来显著的性能提升，标记少量目标样本通常是可行的，而且特别值得。受此启发，我们旨在开发用于医学图像分割的半监督域适应（SSDA），而这在很大程度上尚未得到充分探索。因此，我们建议以统一的方式利用有标签的源域和目标域数据，以及无标签的目标数据。具体而言，我们提出了一种新颖的不对称协同训练（ACT）框架，以整合这些子集，并避免源域数据占据主导地位。遵循分而治之的策略，我们将SSDA中的标签监督明确解耦为两个不对称子任务，包括半监督学习（SSL）和UDA，并利用来自两个分割器的不同知识，以考虑源域和目标域标签监督之间的差异。然后，基于置信度感知伪标签，通过相互迭代教学，将在两个模块中学到的知识与ACT进行自适应整合。此外，通过指数MixUp衰减方案对伪标签噪声进行了很好的控制，以实现平滑传播。使用BraTS18数据库进行的跨模态脑肿瘤MRI分割任务实验表明，即使标记的目标样本有限，ACT相对于UDA和当前最先进的SSDA方法也有显著改进，并接近监督联合训练的“上限”。
