SSMSNet: Scribble-Supervised Myocardial Scar Segmentation in Late Gadolinium Enhancement Images
Diagnostics · 18.06.2026 · 16 (12) · 1895
DOI: 10.3390/diagnostics16121895 ↗Kurz gefasst
Fachartikel in der Zeitschrift Diagnostics, online erschienen am 18.06.2026. Verfasst von Xuewen Liao und 6 weiteren Autor:innen. Beteiligte Einrichtung: Shengli Clinical Medical College of Fujian Medical University, Fuzhou 350013, China. Die Arbeit nennt 31 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Fujian Province Science and Technology Association, Science and Technology Project of Fujian Province.
Abstract (Originalsprache)
Background
Myocardial scar segmentation from late gadolinium enhancement (LGE) cardiac magnetic resonance (CMR) images plays an important role in cardiac disease assessment and prognosis evaluation. However, accurate scar annotation is labor-intensive and requires substantial clinical expertise because scar regions are typically small, irregularly shaped, and characterized by ambiguous boundaries. Although scribble supervision provides a more practical alternative to dense annotation by substantially reducing labeling costs, the extreme sparsity of scribbles and the high similarity between scar tissue and surrounding myocardium make accurate weakly supervised segmentation challenging.
Methods
To address these challenges, we propose SSMSNet, a novel scribble-supervised framework for myocardial scar segmentation. Specifically, a weakly supervised anatomical segmentation network is first employed to provide reliable myocardial structural priors and suppress irrelevant background interference. Subsequently, a local distance prior map is dynamically generated from scribble annotations, and a corresponding loss is introduced to enhance structural awareness and improve training stability. Meanwhile, by leveraging the spatial correlation between the myocardium and scar regions, teacher–student consistency supervision progressively recovers more complete scar structures from sparse annotations. Furthermore, a detail-aware feature enhancement module strengthens low-level representations through contextual interactions and attention mechanisms, improving the perception of scars with ambiguous boundaries.
Results
Extensive experiments conducted on two public cardiac pathology datasets demonstrate that the proposed framework consistently outperforms state-of-the-art scribble-supervised methods and achieves competitive performance compared with fully supervised methods.
Conclusions
The proposed SSMSNet effectively alleviates the limitations imposed by scribble annotations by integrating anatomical guidance, local distance priors, and consistency learning. These findings suggest that the framework provides an effective and annotation-efficient solution for myocardial scar segmentation in LGE CMR images.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Shengli Clinical Medical College of Fujian Medical University, Fuzhou 350013, China.
- Department of Cardiology, Shengli Clinical Medical College of Fujian Medical University, Fuzhou University Affiliated Provincial Hospital, Fujian Provincial Hospital, No. 134 East Street, Gulou District, Fuzhou 350001,…
- College of Physics and Information Engineering, Fuzhou University, Fuzhou 350116, China.
ORCID-Profile: Lin Pan, Mingjing Yang, Jiancheng Zhang
Förderung
- National Natural Science Foundation of China
- Fujian Province Science and Technology Association
- Science and Technology Project of Fujian Province
Zitieren
APA
Liao, X., Yang, K., Lin, X., Pan, L., Lin, Y., Yang, M., & Zhang, J. (2026). SSMSNet: Scribble-Supervised Myocardial Scar Segmentation in Late Gadolinium Enhancement Images. Diagnostics, 16(12), 1895. https://doi.org/10.3390/diagnostics16121895
Vancouver
Liao X, Yang K, Lin X, Pan L, Lin Y, Yang M, et al. SSMSNet: Scribble-Supervised Myocardial Scar Segmentation in Late Gadolinium Enhancement Images. Diagnostics. 2026;16(12):1895. doi:10.3390/diagnostics16121895
BibTeX
@article{Liao2026,
author = {Liao, Xuewen and Yang, Kangwen and Lin, Xingtao and Pan, Lin and Lin, Yazhou and Yang, Mingjing and Zhang, Jiancheng},
title = {SSMSNet: Scribble-Supervised Myocardial Scar Segmentation in Late Gadolinium Enhancement Images},
journal = {Diagnostics},
year = {2026},
volume = {16},
number = {12},
pages = {1895},
doi = {10.3390/diagnostics16121895}
}
EUVIMED