Abstract
Diabetic foot syndrome is one of the chronic diabetic complications that present itself as foot ulceration, which in many cases leads to limb amputation. Moreover, it is linked to the high percentage of post-amputation mortality within the period of five years. Thus it is crucial to diagnose and plan careful treatment properly in the early stages. Diagnosis is often time-consuming and requires a skilled clinician who can differentiate between similarities between diabetes-related and other types of ulcers by evaluating their morphology and location. To mitigate diagnostic issues, we propose an improved version of the nnU-Net architecture with residual short skip connections in the encoder part and additional mixup augmentation as the preprocessing step. The obtained results on the DFUC2022 challenge dataset show that our improvements can boost overall performance for ulcer segmentation tasks, even in scenarios where targeted structures are heterogeneous and under high imbalance conditions in the evaluated dataset. With our approach we achieved 9th place with a Dice score of 0.6975.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Cassidy, B., et al.: The DFUC 2020 dataset: analysis towards diabetic foot ulcer detection. touchREVIEWS Endocrinol. 17, 5–11 (2021). https://doi.org/10.17925/EE.2021.17.1.5, https://www.touchendocrinology.com/diabetes/journal-articles/the-dfuc-2020-dataset-analysis-towards-diabetic-foot-ulcer-detection/1
Drozdzal, M., Vorontsov, E., Chartrand, G., Kadoury, S., Pal, C.: The importance of skip connections in biomedical image segmentation. In: Carneiro, G., et al. (eds.) Deep Learning and Data Labeling for Medical Applications, pp. 179–187. Springer, Cham (2016)
Dutta, A., Zisserman, A.: The via annotation software for images, audio and video. In: Proceedings of the 27th ACM International Conference on Multimedia (2019)
Gazda, M., Bugata, P., Gazda, J., Hubacek, D., Hresko, D.J., Drotar, P.: Mixup augmentation for kidney and kidney tumor segmentation. In: Heller, N., Isensee, F., Trofimova, D., Tejpaul, R., Papanikolopoulos, N., Weight, C. (eds.) Kidney and Kidney Tumor Segmentation, pp. 90–97. Springer, Cham (2022)
Guo, J., Dardik, A., Fang, K., Huang, R., Gu, Y.: Meta-analysis on the treatment of diabetic foot ulcers with autologous stem cells. Stem Cell Res. Ther. 8(1), 1–8 (2017)
Isensee, F., Jaeger, P.F., Kohl, S.A.A., Petersen, J., Maier-Hein, K.H.: NNU-net: a self-configuring method for deep learning-based biomedical image segmentation. Nat. Methods 18(2), 203–211 (2021). https://doi.org/10.1038/s41592-020-01008-z, https://doi.org/10.1038/s41592-020-01008-z
Kendrick, C., et al.: Translating clinical delineation of diabetic foot ulcers into machine interpretable segmentation (2022). https://doi.org/10.48550/ARXIV.2204.11618, https://arxiv.org/abs/2204.11618
Roche, M.M., Wang, P.P.: Factors associated with a diabetes diagnosis and late diabetes diagnosis for males and females. J. Clin. Transl. Endocrinol. 1(3), 77–84 (2014)
Ronneberger, O., Fischer, P., Brox, T.: U-Net: convolutional networks for biomedical image segmentation. In: Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F. (eds.) MICCAI 2015. LNCS, vol. 9351, pp. 234–241. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-24574-4_28
Sun, H., et al.: IDF diabetes atlas: global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes Res. Clin. Pract. 183, 109119 (2022)
Tsai, T.H., Huang, S.A.: Refined u-net: a new semantic technique on hand segmentation. Neurocomputing, 495, 1–10 (2022). https://doi.org/10.1016/j.neucom.2022.04.079, https://www.sciencedirect.com/science/article/pii/S0925231222004696
Yap, M.H., Kendrick, C., Reeves, N.D., Goyal, M., Pappachan, J.M., Cassidy, B.: Development of diabetic foot ulcer datasets: an overview. Diabet. Foot Ulcers Grand Challenge 13183, 1–18 (2021)
Yap, M.H., et al.: Diabetic foot ulcers grand challenge 2022 (2021). https://doi.org/10.5281/zenodo.6389665, https://doi.org/10.5281/zenodo.6389665
Yun, S., Han, D., Chun, S., Oh, S., Yoo, Y., Choe, J.: CutMix: regularization strategy to train strong classifiers with localizable features. In: 2019 IEEE/CVF International Conference on Computer Vision (ICCV), pp. 6022–6031. IEEE Computer Society, Los Alamitos, CA, USA (2019). https://doi.org/10.1109/ICCV.2019.00612, https://doi.ieeecomputersociety.org/10.1109/ICCV.2019.00612
Zhang, H., Cisse, M., Dauphin, Y.N., Lopez-Paz, D.: mixup: beyond empirical risk minimization (2017). https://doi.org/10.48550/ARXIV.1710.09412, https://arxiv.org/abs/1710.09412
Zhou, J., Lu, Y., Tao, S., Cheng, X., Huang, C.: E-RES U-Net: an improved u-net model for segmentation of muscle images. Expert Syst. Appl. 185, 115625 (2021). https://doi.org/10.1016/j.eswa.2021.115625, https://www.sciencedirect.com/science/article/pii/S0957417421010198
Acknowledgements
This work was supported by the Scientific Grant Agency of the Ministry of Education, Science, Research and Sport of the Slovak Republic and the Slovak Academy of Sciences under contract VEGA 1/0327/20.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Hresko, D.J., Vereb, J., Krigovsky, V., Gayova, M., Drotar, P. (2023). Refined Mixup Augmentation for Diabetic Foot Ulcer Segmentation. In: Yap, M.H., Kendrick, C., Cassidy, B. (eds) Diabetic Foot Ulcers Grand Challenge. DFUC 2022. Lecture Notes in Computer Science, vol 13797. Springer, Cham. https://doi.org/10.1007/978-3-031-26354-5_8
Download citation
DOI: https://doi.org/10.1007/978-3-031-26354-5_8
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-26353-8
Online ISBN: 978-3-031-26354-5
eBook Packages: Computer ScienceComputer Science (R0)