IEEE Trans Med Imaging. 2023 Dec;42(12):3919-3931. doi: 10.1109/TMI.2023.3318006. Epub 2023 Nov 30.
Unsupervised domain adaptation (UDA) aims to train a model on a labeled source domain and adapt it to an unlabeled target domain. In medical image segmentation field, most existing UDA methods rely on adversarial learning to address the domain gap between different image modalities. However, this process is complicated and inefficient. In this paper, we propose a simple yet effective UDA method based on both frequency and spatial domain transfer under a multi-teacher distillation framework. In the frequency domain, we introduce non-subsampled contourlet transform for identifying domain-invariant and domain-variant frequency components (DIFs and DVFs) and replace the DVFs of the source domain images with those of the target domain images while keeping the DIFs unchanged to narrow the domain gap. In the spatial domain, we propose a batch momentum update-based histogram matching strategy to minimize the domain-variant image style bias. Additionally, we further propose a dual contrastive learning module at both image and pixel levels to learn structure-related information. Our proposed method outperforms state-of-the-art methods on two cross-modality medical image segmentation datasets (cardiac and abdominal). Codes are avaliable at https://github.com/slliuEric/FSUDA.
无监督领域自适应（UDA）旨在在有标签的源域上训练模型，并将其适应于无标签的目标域。在医学图像分割领域，大多数现有的 UDA 方法依赖于对抗学习来解决不同图像模态之间的领域差距。然而，这个过程复杂且效率低下。在本文中，我们提出了一种简单而有效的 UDA 方法，该方法基于多教师蒸馏框架下的频率和空间域迁移。在频率域中，我们引入非下采样轮廓变换来识别域不变和域变化的频率分量（DIFs 和 DVFs），并用目标域图像的 DVFs 替换源域图像的 DVFs，同时保持 DIFs 不变，以缩小域差距。在空间域中，我们提出了一种基于批量动量更新的直方图匹配策略，以最小化域变化的图像风格偏差。此外，我们还在图像和像素级别进一步提出了双对比学习模块，以学习结构相关信息。我们的方法在两个跨模态医学图像分割数据集（心脏和腹部）上优于最先进的方法。代码可在 https://github.com/slliuEric/FSUDA 上获得。
