TY - JOUR
T1 - Source free domain adaptation for medical image segmentation with fourier style mining
AU - Yang, Chen
AU - Guo, Xiaoqing
AU - Chen, Zhen
AU - Yuan, Yixuan
N1 - This work was supported by Shenzhen-Hong Kong Innovation Circle Category D Project SGDX2019081623300177 (CityU 9240008) and Hong Kong Research Grants Council (RGC) General Research Fund 11211221 (CityU 9043152). We would like to thank Dr. Hu Jiancong from Department of Endoscopic Surgery, the Sixth Affiliated Hospital, Sun Yat-sen University for providing clinical knowledge for endoscopy disease diagnosis.
Publisher Copyright:
© 2022 Elsevier B.V.
PY - 2022/7
Y1 - 2022/7
N2 - Unsupervised domain adaptation (UDA) aims to exploit the knowledge learned from a labeled source dataset to solve similar tasks in a new unlabeled target domain. Existing UDA techniques typically assume that samples from source and target domains are freely accessible during the training. However, it may be impractical to access source images due to privacy concerns, especially in medical imaging scenarios with the patient information. To tackle this issue, we devise a novel source free domain adaptation framework with fourier style mining, where only a well-trained source segmentation model is available for the adaptation to the target domain. Our framework is composed of two stages: a generation stage and an adaptation stage. In the generation stage, we design a Fourier Style Mining (FSM) generator to inverse source-like images through statistic information of the pretrained source model and mutual Fourier Transform. These generated source-like images can provide source data distribution and benefit the domain alignment. In the adaptation stage, we design a Contrastive Domain Distillation (CDD) module to achieve feature-level adaptation, including a domain distillation loss to transfer relation knowledge and a domain contrastive loss to narrow down the domain gap by a self-supervised paradigm. Besides, a Compact-Aware Domain Consistency (CADC) module is proposed to enhance consistency learning by filtering out noisy pseudo labels with shape compactness metric, thus achieving output-level adaptation. Extensive experiments on cross-device and cross-centre datasets are conducted for polyp and prostate segmentation, and our method delivers impressive performance compared with state-of-the-art domain adaptation methods.
AB - Unsupervised domain adaptation (UDA) aims to exploit the knowledge learned from a labeled source dataset to solve similar tasks in a new unlabeled target domain. Existing UDA techniques typically assume that samples from source and target domains are freely accessible during the training. However, it may be impractical to access source images due to privacy concerns, especially in medical imaging scenarios with the patient information. To tackle this issue, we devise a novel source free domain adaptation framework with fourier style mining, where only a well-trained source segmentation model is available for the adaptation to the target domain. Our framework is composed of two stages: a generation stage and an adaptation stage. In the generation stage, we design a Fourier Style Mining (FSM) generator to inverse source-like images through statistic information of the pretrained source model and mutual Fourier Transform. These generated source-like images can provide source data distribution and benefit the domain alignment. In the adaptation stage, we design a Contrastive Domain Distillation (CDD) module to achieve feature-level adaptation, including a domain distillation loss to transfer relation knowledge and a domain contrastive loss to narrow down the domain gap by a self-supervised paradigm. Besides, a Compact-Aware Domain Consistency (CADC) module is proposed to enhance consistency learning by filtering out noisy pseudo labels with shape compactness metric, thus achieving output-level adaptation. Extensive experiments on cross-device and cross-centre datasets are conducted for polyp and prostate segmentation, and our method delivers impressive performance compared with state-of-the-art domain adaptation methods.
KW - Source Free Domain Adaptation
KW - Fourier Style Mining
KW - Contrastive Domain Distillation
KW - Consistency Learning
UR - http://www.scopus.com/inward/record.url?scp=85129547297&partnerID=8YFLogxK
U2 - 10.1016/j.media.2022.102457
DO - 10.1016/j.media.2022.102457
M3 - Journal article
C2 - 35461016
AN - SCOPUS:85129547297
SN - 1361-8415
VL - 79
JO - Medical Image Analysis
JF - Medical Image Analysis
M1 - 102457
ER -