MSCDA: Multi-level semantic-guided contrast improves unsupervised domain adaptation for breast MRI segmentation in small datasets
Kuang, Sheng; Woodruff, Henry C.; Granzier, Renee; van Nijnatten, Thiemo J.A.; Lobbes, Marc B.I.; Smidt, Marjolein L.; Lambin, Philippe; Mehrkanoon, Siamak
(2023) Neural Networks, volume 165, pp. 119 - 134
(Article)
Abstract
Deep learning (DL) applied to breast tissue segmentation in magnetic resonance imaging (MRI) has received increased attention in the last decade, however, the domain shift which arises from different vendors, acquisition protocols, and biological heterogeneity, remains an important but challenging obstacle on the path towards clinical implementation. In this paper,
... read more
we propose a novel Multi-level Semantic-guided Contrastive Domain Adaptation (MSCDA) framework to address this issue in an unsupervised manner. Our approach incorporates self-training with contrastive learning to align feature representations between domains. In particular, we extend the contrastive loss by incorporating pixel-to-pixel, pixel-to-centroid, and centroid-to-centroid contrasts to better exploit the underlying semantic information of the image at different levels. To resolve the data imbalance problem, we utilize a category-wise cross-domain sampling strategy to sample anchors from target images and build a hybrid memory bank to store samples from source images. We have validated MSCDA with a challenging task of cross-domain breast MRI segmentation between datasets of healthy volunteers and invasive breast cancer patients. Extensive experiments show that MSCDA effectively improves the model's feature alignment capabilities between domains, outperforming state-of-the-art methods. Furthermore, the framework is shown to be label-efficient, achieving good performance with a smaller source dataset. The code is publicly available at https://github.com/ShengKuangCN/MSCDA.
show less
Download/Full Text
Keywords: Breast segmentation, Contrastive learning, Unsupervised domain adaptation, Cognitive Neuroscience, Artificial Intelligence
ISSN: 0893-6080
Publisher: Elsevier Limited
Note: Funding Information: The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Authors acknowledge financial support from ERC advanced grant (ERC-ADG-2015 n° 694812 - Hypoximmuno), ERC-2020-PoC: 957565-AUTO.DISTINCT. Authors also acknowledge financial support from the European Union's Horizon 2020 research and innovation programme under grant agreement: ImmunoSABR n° 733008, CHAIMELEON n° 952172, EuCanImage n° 952103. This work was also partially supported by the Dutch Cancer Society (KWF Kankerbestrijding), The Netherlands, project number 14449. Funding Information: The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Authors acknowledge financial support from ERC advanced grant ( ERC-ADG-2015 n° 694812 - Hypoximmuno), ERC-2020-PoC: 957565-AUTO.DISTINCT . Authors also acknowledge financial support from the European Union’s Horizon 2020 research and innovation programme under grant agreement: ImmunoSABR n° 733008 , CHAIMELEON n° 952172 , EuCanImage n° 952103 . This work was also partially supported by the Dutch Cancer Society (KWF Kankerbestrijding), The Netherlands , project number 14449 . Publisher Copyright: © 2023 The Author(s)
(Peer reviewed)
See more statistics about this item