Abstract
Adapting a medical image segmentation model to a new domain is important for improving its cross-domain transferability, and due to the expensive annotation process, Unsupervised Domain Adaptation (UDA) is appealing where only unlabeled images are needed for the adaptation. Existing UDA methods are mainly based on image or feature alignment with adversarial training for regularization, and they are limited by insufficient supervision in the target domain. In this paper, we propose an enhanced Filtered Pseudo Label (FPL+)-based UDA method for 3D medical image segmentation. It first uses cross-domain data augmentation to translate labeled images in the source domain to a dual-domain training set consisting of a pseudo source-domain set and a pseudo target-domain set. To leverage the dual-domain augmented images to train a pseudo label generator, domain-specific batch normalization layers are used to deal with the domain shift while learning the domain-invariant structure features, generating high-quality pseudo labels for target-domain images. We then combine labeled source-domain images and target-domain images with pseudo labels to train a final segmentor, where image-level weighting based on uncertainty estimation and pixel-level weighting based on dual-domain consensus are proposed to mitigate the adverse effect of noisy pseudo labels. Experiments on three public multi-modal datasets for Vestibular Schwannoma, brain tumor and whole heart segmentation show that our method surpassed ten state-of-the-art UDA methods, and it even achieved better results than fully supervised learning in the target domain in some cases.
Abstract (translated)
将医学图像分割模型适应新领域非常重要，以提高其跨领域可迁移性。由于昂贵的注释过程，无监督域自适应（UDA）方法在仅需要未标注图像的适应性方面具有吸引力。现有的UDA方法主要基于图像或特征与对抗训练进行对齐来解决正则化问题，但它们在目标域中的监督不足。在本文中，我们提出了一个增强的带滤波伪标签（FPL+）-based UDA方法，用于3D医学图像分割。它首先使用跨域数据增强将源域中的带标签图像转换为由伪源域集和伪目标域集组成的双域训练集。为了利用双域增强的图像来训练伪标签生成器，我们使用领域特定的批归一化层来处理领域漂移，同时学习目标域不变的结构特征，为目标域图像生成高质量伪标签。然后，我们将带标签源域图像和目标域图像与伪标签结合，用于训练最终分割器。我们提出基于不确定度估计的图像级加权和基于双域共识的像素级加权来减轻噪声伪标签的负面影响。在三个公开的多模态数据集（Vestibular Schwannoma、脑肿瘤和整个心脏分割）上进行实验，我们的方法超越了10个最先进的UDA方法，在某些情况下，甚至取得了比完全监督学习在目标域中更好的结果。
URL
https://arxiv.org/abs/2404.04971
