DDA-Net: Unsupervised cross-modality medical image segmentation via dual domain adaptation

• Published in: Computer methods and programs in biomedicine Vol. 213; p. 106531
• Main Authors: Bian, Xuesheng, Luo, Xiongbiao, Wang, Cheng, Liu, Weiquan, Lin, Xiuhong
• Format: Journal Article
• Language: English
• Published: Ireland Elsevier B.V 01-01-2022
• Subjects: Brain - diagnostic imaging; Cross-modality; Domain adaptation; Heart; Image Processing, Computer-Assisted; Magnetic Resonance Imaging; Medical image; Segmentation; Unsupervised learning; Medical image; Domain adaptation; Cross-modality; Segmentation; Unsupervised learning
• ISSN: 0169-2607; 1872-7565
• DOI: 10.1016/j.cmpb.2021.106531

•We propose DDA-Net, a novel domain adaptation method, for ten categories of complicated unsupervised pixel-wise semantic segmentation, which performs domain adaptation in both feature-space and image-space.•Using a cross-modality auto-encoder, DDA-Net maps cross-modality medical images into a feature shared subspace, and effectively releases the structural distortion caused by DCNs trained with insufficient data.•Experiments demonstrate that DDA-Net with dual domain adaptation effectively improves the accuracy for unsupervised segmentation and achieves state-of-the-art performance in cross-modality head and heart image segmentation.
[Display omitted]
Background and Objective: Deep convolutional networks are powerful tools for single-modality medical image segmentation, whereas generally require semantic labelling or annotation that is laborious and time-consuming. However, domain shift among various modalities critically deteriorates the performance of deep convolutional networks if only trained by single-modality labelling data.
Methods: In this paper, we propose an end-to-end unsupervised cross-modality segmentation network, DDA-Net, for accurate medical image segmentation without semantic annotation or labelling on the target domain. To close the domain gap, different images with domain shift are mapped into a shared domain-invariant representation space. In addition, spatial position information, which benefits the spatial structure consistency for semantic information, is preserved by an introduced cross-modality auto-encoder.
Results: We validated the proposed DDA-Net method on cross-modality medical image datasets of brain images and heart images. The experimental results show that DDA-Net effectively alleviates domain shift and suppresses model degradation.
Conclusions: The proposed DDA-Net successfully closes the domain gap between different modalities of medical image, and achieves state-of-the-art performance in cross-modality medical image segmentation. It also can be generalized for other semi-supervised or unsupervised segmentation tasks in some other field.