TY - GEN
T1 - Domain Specific Convolution and High Frequency Reconstruction Based Unsupervised Domain Adaptation for Medical Image Segmentation
AU - Hu, Shishuai
AU - Liao, Zehui
AU - Xia, Yong
N1 - Publisher Copyright:
© 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2022
Y1 - 2022
N2 - Although deep learning models have achieved remarkable success in medical image segmentation, the domain shift issue caused mainly by the highly variable quality of medical images is a major hurdle that prevents these models from being deployed for real clinical practices, since no one can predict the performance of a ‘well-trained’ model on a set of unseen clinical data. Previously, many methods have been proposed based on, for instance, CycleGAN or the Fourier transform to address this issue, which, however, suffer from either an inadequate ability to preserve anatomical structures or unexpectedly introduced artifacts. In this paper, we propose a multi-source-domain unsupervised domain adaptation (UDA) method called Domain specific Convolution and high frequency Reconstruction (DoCR) for medical image segmentation. We design an auxiliary high frequency reconstruction (HFR) task to facilitate UDA, and hence avoid the interference of the artifacts generated by the low-frequency component replacement. We also construct the domain specific convolution (DSC) module to boost the segmentation model’s ability to domain-invariant features extraction. We evaluate DoCR on a benchmark fundus image dataset. Our results indicate that the proposed DoCR achieves superior performance over other UDA methods in multi-domain joint optic cup and optic disc segmentation. Code is available at: https://github.com/ShishuaiHu/DoCR.
AB - Although deep learning models have achieved remarkable success in medical image segmentation, the domain shift issue caused mainly by the highly variable quality of medical images is a major hurdle that prevents these models from being deployed for real clinical practices, since no one can predict the performance of a ‘well-trained’ model on a set of unseen clinical data. Previously, many methods have been proposed based on, for instance, CycleGAN or the Fourier transform to address this issue, which, however, suffer from either an inadequate ability to preserve anatomical structures or unexpectedly introduced artifacts. In this paper, we propose a multi-source-domain unsupervised domain adaptation (UDA) method called Domain specific Convolution and high frequency Reconstruction (DoCR) for medical image segmentation. We design an auxiliary high frequency reconstruction (HFR) task to facilitate UDA, and hence avoid the interference of the artifacts generated by the low-frequency component replacement. We also construct the domain specific convolution (DSC) module to boost the segmentation model’s ability to domain-invariant features extraction. We evaluate DoCR on a benchmark fundus image dataset. Our results indicate that the proposed DoCR achieves superior performance over other UDA methods in multi-domain joint optic cup and optic disc segmentation. Code is available at: https://github.com/ShishuaiHu/DoCR.
KW - Domain specific convolution
KW - High frequency reconstruction
KW - Medical image segmentation
KW - Unsupervised domain adaptation
UR - http://www.scopus.com/inward/record.url?scp=85139009047&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-16449-1_62
DO - 10.1007/978-3-031-16449-1_62
M3 - 会议稿件
AN - SCOPUS:85139009047
SN - 9783031164484
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 650
EP - 659
BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference, Proceedings
A2 - Wang, Linwei
A2 - Dou, Qi
A2 - Fletcher, P. Thomas
A2 - Speidel, Stefanie
A2 - Li, Shuo
PB - Springer Science and Business Media Deutschland GmbH
T2 - 25th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2022
Y2 - 18 September 2022 through 22 September 2022
ER -