EUVIMEDEUVIMED entdecken
Crossref

Topology Optimization in Medical Image Segmentation With Fast χ Euler Characteristic

Liu Li, Qiang Ma, Cheng Ouyang, Johannes C Paetzold, Daniel Rueckert, Bernhard Kainz

IEEE Transactions on Medical Imaging · 2025-12 · 44 (12) · 5221-5232

DOI: 10.1109/tmi.2025.3589495 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Medical Imaging, 2025-12. Verfasst von Liu Li und 5 weiteren Autor:innen. Hauptthemen laut MEDLINE: Image Processing, Computer-Assisted, Deep Learning. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 51 Literaturangaben. Gefördert von: ERC–Project Medical Image Analysis with Normative Machine Learning, ERC–Project Deep Learning for Medical Imaging: Learning Clinically Useful Information from Images, ERC–Project Developing Human Connectome Project (dHCP) Data Collection.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Deep learning-based medical image segmentation techniques have shown promising results when evaluated based on conventional metrics such as the Dice score or Intersection-over-Union. However, these fully automatic methods often fail to meet clinically acceptable accuracy, especially when topological constraints should be observed, e.g., continuous boundaries or closed surfaces. In medical image segmentation, the correctness of a segmentation in terms of the required topological genus sometimes is even more important than the pixel-wise accuracy. Existing topology-aware approaches commonly estimate and constrain the topological structure via the concept of persistent homology (PH). However, these methods are difficult to implement for high dimensional data due to their polynomial computational complexity. To overcome this problem, we propose a novel and fast approach for topology-aware segmentation based on the Euler Characteristic ( $\chi $ ). First, we propose a fast formulation for $\chi $ computation in both 2D and 3D. The scalar $\chi $ error between the prediction and ground-truth serves as the topological evaluation metric. Then we estimate the spatial topology correctness of any segmentation network via a so-called topological violation map, i.e., a detailed map that highlights regions with $\chi $ errors. Finally, the segmentation results from the arbitrary network are refined based on the topological violation maps by a topology-aware correction network. Our experiments are conducted on both 2D and 3D datasets and show that our method can significantly improve topological correctness while preserving pixel-wise segmentation accuracy.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Image Processing, Computer-AssistedDeep LearningHumansAlgorithms

Förderung

  • ERC–Project Medical Image Analysis with Normative Machine Learning
  • ERC–Project Deep Learning for Medical Imaging: Learning Clinically Useful Information from Images
  • ERC–Project Developing Human Connectome Project (dHCP) Data Collection
  • High-Tech Agenda Bavaria and HPC resources provided by the Erlangen National High Performance Computing Center
  • NHR funding is provided by federal and Bavarian state authorities
  • Deutsche Forschungsgemeinschaft

Zitieren

APA

Li, L., Ma, Q., Ouyang, C., Paetzold, J. C., Rueckert, D., & Kainz, B. (2025). Topology Optimization in Medical Image Segmentation With Fast χ Euler Characteristic. IEEE Transactions on Medical Imaging, 44(12), 5221-5232. https://doi.org/10.1109/tmi.2025.3589495

Vancouver

Li L, Ma Q, Ouyang C, Paetzold JC, Rueckert D, Kainz B. Topology Optimization in Medical Image Segmentation With Fast χ Euler Characteristic. IEEE Transactions on Medical Imaging. 2025;44(12):5221-5232. doi:10.1109/tmi.2025.3589495

BibTeX
@article{Li2025,
  author = {Li, Liu and Ma, Qiang and Ouyang, Cheng and Paetzold, Johannes C and Rueckert, Daniel and Kainz, Bernhard},
  title = {Topology Optimization in Medical Image Segmentation With Fast χ Euler Characteristic},
  journal = {IEEE Transactions on Medical Imaging},
  year = {2025},
  volume = {44},
  number = {12},
  pages = {5221--5232},
  doi = {10.1109/tmi.2025.3589495}
}
Kostenlos starten →