EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Dynamic thresholding and robust contrastive techniques for enhanced semi-supervised cardiac segmentation

Yafei Mi, Jie Zhang, Hui Jin, Jiayi Yin, Yuyang He, Guojie Xie, Shiqi Yan, Yongping Fu iD

PLOS One · 06.04.2026 · 21 (4) · e0342567

DOI: 10.1371/journal.pone.0342567 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS One, online erschienen am 06.04.2026. Verfasst von Yafei Mi und 7 weiteren Autor:innen. Beteiligte Einrichtung: Department of Cardiology, Taizhou Hospital of Zhejiang Province affiliated to Wenzhou Medical University, Linhai, Zhejiang, China. Hauptthemen laut MEDLINE: Heart, Image Processing, Computer-Assisted. Die Arbeit nennt 66 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Cardiac segmentation plays a crucial role in the diagnosis of cardiovascular diseases. However, the manual annotation of cardiac structures is a labor-intensive and time-consuming task that requires highly trained experts. Moreover, the availability of labeled data for training segmentation models is often limited due to the challenges associated with acquiring accurate annotations. To address this issue, we propose a novel semi-supervised cardiac segmentation framework that only needs a small set of labeled data with a larger pool of unlabeled data. We propose three strategies: dynamic pseudo-label threshold map, robust entropy minimization and contrastive consistency from the perspective of pseudo-labeling, entropy minimization and consistency regularization. Specifically, we propose a pixel-wise, class-wise and adaptive map to generate threshold maps and use the map for robust entropy minimization to reduce the noise from low-confidence samples. Besides, to utilize the unlabeled data sufficiently, we add contrastive consistency loss to implement regularization. Extensive experiments on the ACDC and MMWHS datasets demonstrate that our method achieves competitive performance compared to state-of-the-art approaches across various labeled data ratios. Ablation studies further validate the effectiveness and robustness of each component. Our framework shows strong potential for accurate diagnosis with limited annotations, and our code will be made publicly available.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

HeartImage Processing, Computer-AssistedHumansAlgorithmsEntropy

Autor:innen und Institutionen

  • Department of Cardiology, Taizhou Hospital of Zhejiang Province affiliated to Wenzhou Medical University, Linhai, Zhejiang, China.
  • Cardiovascular Disease Clinical Translational Innovation Center of Shaoxing University, Shaoxing, Zhejiang, China.
  • Innovation Centre for Information, Binjiang Institute of Zhejiang University, Hangzhou, Zhejiang, China.
  • Department of Cardiology, Linhai Second People's Hospital, Linhai, Zhejiang, China.
  • Digital Zhejiang Technology Operation Co., Ltd., Hangzhou, Zhejiang, China.
  • Department of Cardiovascular Medicine, Affiliated Hospital of Shaoxing University, Shaoxing, Zhejiang, China.

ORCID-Profile: Yongping Fu

Zitieren

APA

Mi, Y., Zhang, J., Jin, H., Yin, J., He, Y., Xie, G., Yan, S., & Fu, Y. (2026). Dynamic thresholding and robust contrastive techniques for enhanced semi-supervised cardiac segmentation. PLOS One, 21(4), e0342567. https://doi.org/10.1371/journal.pone.0342567

Vancouver

Mi Y, Zhang J, Jin H, Yin J, He Y, Xie G, et al. Dynamic thresholding and robust contrastive techniques for enhanced semi-supervised cardiac segmentation. PLOS One. 2026;21(4):e0342567. doi:10.1371/journal.pone.0342567

BibTeX
@article{Mi2026,
  author = {Mi, Yafei and Zhang, Jie and Jin, Hui and Yin, Jiayi and He, Yuyang and Xie, Guojie and Yan, Shiqi and Fu, Yongping},
  title = {Dynamic thresholding and robust contrastive techniques for enhanced semi-supervised cardiac segmentation},
  journal = {PLOS One},
  year = {2026},
  volume = {21},
  number = {4},
  pages = {e0342567},
  doi = {10.1371/journal.pone.0342567}
}
Kostenlos starten →