Abstract
Image segmentation is a complex task that aims to simultaneously satisfy various quality criteria. In this context, topology is being increasingly considered. Guaranteeing correct topological properties is indeed crucial for objects presenting challenging shapes. Designing topology-aware metrics is then relevant, both for assessing the quality of segmentation results and for designing losses involved in learning procedures. In this article, we introduce ccDice (connected component Dice), a topological metric that generalises the popular Dice score. By contrast to Dice, that acts at the scale of pixels, ccDice acts at the scale of connected components of the compared objects, leading to a topological assessment of their relative structure and embedding. ccDice is a simple, explainable, normalized and low-computational topological metric. We provide a formal definition of ccDice, an algorithmic scheme for computing it, and we assess its behaviour by comparison to other usual topological metrics. Code is available on GitHub: https://github.com/PierreRouge/ccDice.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Bauer, U., Lesnick, M.: Induced matchings and the algebraic stability of persistence barcodes. J. Comput. Geom. 6, 162–191 (2015)
Byrne, N., et al.: A persistent homology-based topological loss for CNN-based multiclass segmentation of CMR. IEEE Trans. Med. Imaging 42, 3–14 (2022)
Carlinet, E., Géraud, T.: A comparative review of component tree computation algorithms. IEEE Trans. Image Process. 23, 3885–3895 (2014)
Clough, J.R., et al.: A topological loss function for deep-learning based image segmentation using persistent homology. IEEE Trans. Pattern Anal. 44, 8766–8778 (2020)
Couprie, M., Bertrand, G.: New characterizations of simple points in 2D, 3D, and 4D discrete spaces. IEEE Trans. Pattern Anal. 31, 637–648 (2009)
Dice, L.R.: Measures of the amount of ecologic association between species. Ecology 26, 297–302 (1945)
Edelsbrunner, H., Harrer, J.: Persistent homology - a survey. Contemp. Math. 453, 257–282 (2008)
Fraz, M.M., et al.: An ensemble classification-based approach applied to retinal blood vessel segmentation. IEEE Trans. Bio. Med. Eng. 59, 2538–2548 (2012)
Gupta, S., et al.: Topology-aware uncertainty for image segmentation. In: NeurIPS, Procs. (2024)
Hu, X., et al.: Topology-preserving deep image segmentation. In: NeurIPS, Procs. (2019)
Hu, X., et al.: Topology-aware segmentation using discrete Morse theory. In: ICLR, Procs. (2021)
Kovalevsky, V.A.: Finite topology as applied to image analysis. Comput. Vision. Graph. 46, 141–161 (1989)
Mazo, L., et al.: Paths, homotopy and reduction in digital images. Acta Appl. Math. 113, 167–193 (2011)
Mazo, L., et al.: Digital imaging: a unified topological framework. J. Math. Imaging Vis. 44, 19–37 (2012)
Passat, N., Mendes Forte, J., Kenmochi, Y.: Morphological hierarchies: a unifying framework with new trees. J. Math. Imaging Vis. 65, 718–753 (2023)
Perret, B., Cousty, J.: Component tree loss function: Definition and optimization. In: DGMM, Procs., pp. 248–260 (2022)
Rosenfeld, A.: Adjacency in digital pictures. Inform. Control 26, 24–33 (1974)
Rosenfeld, A.: Digital topology. Am. Math. Mon. 86, 621–630 (1979)
Saha, P.K., Strand, R., Borgefors, G.: Digital topology and geometry in medical imaging: a survey. IEEE Trans. Med. Imaging 34, 1940–1964 (2015)
Salembier, P., Oliveras, A., Garrido, L.: Anti-extensive connected operators for image and sequence processing. IEEE Trans. Image Process. 7, 555–570 (1998)
Shit, S., et al.: clDice–A novel topology-preserving loss function for tubular structure segmentation. In: CVPR, Procs., pp. 16560–16569 (2021)
Stucki, N., et al.: Topologically faithful image segmentation via induced matching of persistence barcodes. In: ICML, Procs., pp. 32698–32727 (2023)
Acknowledgments
This work was supported by the Agence Nationale de la Recherche (Grants ANR-20-CE45-0011, ANR-22-CE45-0034, ANR-22-CE45-0018 and ANR-23-CE45-0015).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Ethics declarations
Disclosure of Interests
The authors have no competing interests to declare that are relevant to the content of this article.
1 Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
Copyright information
© 2025 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Rougé, P., Merveille, O., Passat, N. (2025). ccDice: A Topology-Aware Dice Score Based on Connected Components. In: Chen, C., Singh, Y., Hu, X. (eds) Topology- and Graph-Informed Imaging Informatics. TGI3 2024. Lecture Notes in Computer Science, vol 15239. Springer, Cham. https://doi.org/10.1007/978-3-031-73967-5_2
Download citation
DOI: https://doi.org/10.1007/978-3-031-73967-5_2
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-73966-8
Online ISBN: 978-3-031-73967-5
eBook Packages: Computer ScienceComputer Science (R0)
