Abstract
Single-source domain generalization aims to enhance model performance on unseen target domain test sets using only a single source domain dataset, typically by mitigating domain shifts between domains. In retinal vessel segmentation tasks, differences in dataset composition, such as variations in the proportions of different diseases and imaging noise levels, are considered significant sources of domain shift. However, few previous studies have delved into the mechanisms through which this type of domain shift influences model performance. In this study, we hypothesize that disparities in dataset composition could manifest as differences in distribution patterns of frequency domain features, rendering the model susceptible to overfitting specific patterns. Building on this hypothesis, we propose a novel Frequency Dropout based Single Source Domain Generalization (FD-SDG) framework that employs a Frequency Dropout Randomization mechanism to disentangle complex co-adaptive relationships among features from different frequency bands, thereby enhancing the model’s robustness to variable frequency domain noise patterns in the sample space. Additionally, we introduce a Salient Structure Representation Normalization mechanism to align post-perturbation data features in the feature space using invariant anatomical structures. Through comparison experiments and ablation studies conducted on multiple sets of fundus images across-domain experiments, our method achieves state-of-the-art performance, underscoring its high generalizability and robustness.
B. Li and H. Li—Contributing equally to this work
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Hu, S., Liao, Z., Zhang, J., Xia, Y.: Domain and content adaptive convolution based multi-source domain generalization for medical image segmentation. IEEE Trans. Med. Imaging 42, 233–244 (2021)
Li, H., Li, H., Shu, H., Chen, J., Hu, Y., Liu, J.: Self-supervision boosted retinal vessel segmentation for cross-domain data. In: 2023 IEEE 20th International Symposium on Biomedical Imaging (ISBI), pp.1–5 (2023)
Li, H., et al.: RaffeSDG: Random Frequency Filtering enabled Single-source Domain Generalization for Medical Image Segmentation (2024)
Ouyang, C., et al.: Causality-inspired single-source domain generalization for medical image segmentation. IEEE Trans. Med. Imaging 42, 1095–1106 (2021)
Su, Z., Yao, K., Yang, X., Wang, Q., Sun, J., Huang, K.: Rethinking data augmentation for single-source domain generalization in medical image segmentation. In: AAAI Conference on Artificial Intelligence (2022)
Li, H., et al.: A Generic fundus image enhancement network boosted by frequency self-supervised representation learning. Med. Image Anal. 90, 102945 (2023)
Li, H., et al.: Enhancing and adapting in the clinic: source-free unsupervised domain adaptation for medical image enhancement. IEEE Trans. Med. Imaging 43, 1323–1336 (2023)
Li, H., et al.: Frequency-mixed Single-source Domain Generalization for Medical Image Segmentation. ArXiv, abs/2307.09005 (2023)
Xu, Q., Zhang, R., Zhang, Y., Wang, Y., Tian, Q.: A fourier-based framework for domain generalization. IEEE/CVF Conf. Comput. Vis. Pattern Recogn. (CVPR) 2021, 14378–14387 (2021)
Çugu, I., Mancini, M., Chen, Y., Akata, Z.: Attention consistency on visual corruptions for single-source domain generalization. IEEE/CVF Conf. Comput. Vis. Pattern Recogn. Workshops (CVPRW) 2022, 4164–4173 (2022)
Li, H., Li, H., Qiu, Z., Hu, Y., Liu, J.: Domain Adaptive Retinal Vessel Segmentation Guided by High-frequency Component. OMIA@MICCAI (2022)
Hinton, G.E., Srivastava, N., Krizhevsky, A., Sutskever, I., Salakhutdinov, R.. Improving neural networks by preventing co-adaptation of feature detectors. ArXiv, abs/1207.0580 (2012)
Ou, M.,et al.: MVD-Net: Semantic segmentation of cataract surgery using multi-view learning. In: 2022 44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), pp. 5035–5038 (2022)
Srivastava, N., Hinton, G.E., Krizhevsky, A., Sutskever, I., Salakhutdinov, R.: Dropout: a simple way to prevent neural networks from overfitting. J. Mach. Learn. Res. 15, 1929–1958 (2014)
Inoue, H.: Multi-Sample Dropout for Accelerated Training and Better Generalization. ArXiv, abs/1905.09788 (2019)
Vanmali, A.V., Kataria, T., Kelkar, S.G., Gadre, V.M.: Ringing artifacts in wavelet based image fusion: analysis, measurement and remedies. Inf. Fusion 56, 39–69 (2020)
Zhao, L., Wang, L.: Task-specific inconsistency alignment for domain adaptive object detection. IEEE/CVF Conf. Comput. Vis. Pattern Recogn. (CVPR) 2022, 14197–14206 (2022)
Woo, S., Park, J., Lee, J., Kweon, I.: CBAM: Convolutional Block Attention Module. ArXiv, abs/1807.06521 (2018)
Staal, J., Abrà moff, M.D., Niemeijer, M., Viergever, M.A., Ginneken, B.V.: Ridge-based vessel segmentation in color images of the retina. IEEE Trans. Med. Imaging 23, 501–509 (2004)
Gulshan, V., et al.: Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. JAMA 316(22), 2402–2410 (2016)
Zhang, J., Dashtbozorg, B., Bekkers, E.J., Pluim, J.P., Duits, R., Romeny, B.M.: Robust retinal vessel segmentation via locally adaptive derivative frames in orientation scores. IEEE Trans. Med. Imaging 35, 2631–2644 (2016)
Orlando, J.I., Breda, J.B., Keer, K.V., Blaschko, M.B., Blanco, P.J., Bulant, C.A. Towards a glaucoma risk index based on simulated hemodynamics from fundus images. In: International Conference on Medical Image Computing and Computer-Assisted Intervention (2018)
Owen, C.G., et al.: Retinal arteriolar tortuosity and cardiovascular risk factors in a multi-ethnic population study of 10-year-old children; the child heart and health study in England (CHASE). Arterioscler. Thromb. Vasc. Biol. 31, 1933–1938 (2011)
Peng, D., Lei, Y., Hayat, M., Guo, Y., Li, W.: Semantic-aware domain generalized segmentation. IEEE/CVF Conf. Comput. Vis. Pattern Recogn. (CVPR) 2022, 2584–2595 (2022)
Liu, Q., Chen, C., Qin, J., Dou, Q., Heng, P.: FedDG: federated domain generalization on medical image segmentation via episodic learning in continuous frequency space. IEEE/CVF Conf. Comput. Vis. Pattern Recogn. (CVPR) 2021, 1013–1023 (2021)
Acknowledgments
This work was supported in part by General Program of National Natural Science Foundation of China (Grant No. 82272086), Shenzhen Science and Technology Program (JCYJ20210324103800001, JCYJ20220530112609022), Guangdong Basic and Applied Basic Research Fund (2022A1515010487), Nanshan District Healthcare Program (NSZD2023058) and SUSTech Undergraduate Innovation and Entrepreneurship (S202314325015). We appreciate Shenzhen DE Sci&Tech Company for their data support in this study.
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2024 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Li, B. et al. (2024). FD-SDG: Frequency Dropout Based Single Source Domain Generalization Framework for Retinal Vessel Segmentation. In: Huang, DS., Zhang, Q., Guo, J. (eds) Advanced Intelligent Computing in Bioinformatics. ICIC 2024. Lecture Notes in Computer Science(), vol 14881. Springer, Singapore. https://doi.org/10.1007/978-981-97-5689-6_34
Download citation
DOI: https://doi.org/10.1007/978-981-97-5689-6_34
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-97-5688-9
Online ISBN: 978-981-97-5689-6
eBook Packages: Computer ScienceComputer Science (R0)