EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

PM-DUnet: Fusing long-range dependencies and attention in a dual-U architecture for thyroid nodule segmentation

Shaoqiang Wang, Linhao Zhang, Guiling Shi, Zhongran Liu, Yuanyuan Zhang, Tiyao Liu, Yawu Zhao, Yuchen Wang, Xiaochun Cheng iD

PLOS One · 30.07.2026 · 21 (7) · e0353684

DOI: 10.1371/journal.pone.0353684 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS One, online erschienen am 30.07.2026. Verfasst von Shaoqiang Wang und 8 weiteren Autor:innen. Beteiligte Einrichtung: Qingdao University of Technology, Qingdao, Shandong, China. Hauptthemen laut MEDLINE: Thyroid Nodule, Image Processing, Computer-Assisted. Die Arbeit nennt 28 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Doctoral Training Centre at Swansea University.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

For medical image segmentation, accurately balancing local details and global long-range dependencies is critical to tackling thyroid nodule challenges (variable sizes, ambiguous boundaries, complex context). Traditional CNNs excel at local feature extraction but are constrained by local receptive fields, hindering efficient global dependency modeling. To address this, we propose a Parallel Mamba Dual-U Network (PM-DUNet). It adopts a cascaded dual U-Net encoder-decoder for two-stage “coarse-to-fine” segmentation refinement. We design a Multi-Path Parallel Mamba (MPM) module—using State Space Models (SSMs)—to efficiently model global context with linear complexity. Additionally, Squeeze-Excitation Downsampling (SED) and Spatial Attention Upsampling (SAU) modules are integrated to adaptively enhance key features in encoding/decoding. Results show PM-DUNet achieves highly competitive performance and outperforms state-of-the-art methods on most core metrics, verifying its effectiveness and robustness for complex medical image segmentation. Our code is available on https://github.com/Andrevict/MPDUNet .

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Thyroid NoduleImage Processing, Computer-AssistedHumansAlgorithmsConvolutional Neural Networks

Autor:innen und Institutionen

  • Qingdao University of Technology, Qingdao, Shandong, China.
  • China University of Petroleum (East China), Qingdao, China.
  • School of Medical Informational Engineering, Shandong University of Traditional Chinese Medicine, Jinan, Shandong, China.
  • Swansea University, Swansea, Wales, United Kingdom.

ORCID-Profile: Xiaochun Cheng

Förderung

  • Doctoral Training Centre at Swansea University

Zitieren

APA

Wang, S., Zhang, L., Shi, G., Liu, Z., Zhang, Y., Liu, T., Zhao, Y., Wang, Y., & Cheng, X. (2026). PM-DUnet: Fusing long-range dependencies and attention in a dual-U architecture for thyroid nodule segmentation. PLOS One, 21(7), e0353684. https://doi.org/10.1371/journal.pone.0353684

Vancouver

Wang S, Zhang L, Shi G, Liu Z, Zhang Y, Liu T, et al. PM-DUnet: Fusing long-range dependencies and attention in a dual-U architecture for thyroid nodule segmentation. PLOS One. 2026;21(7):e0353684. doi:10.1371/journal.pone.0353684

BibTeX
@article{Wang2026,
  author = {Wang, Shaoqiang and Zhang, Linhao and Shi, Guiling and Liu, Zhongran and Zhang, Yuanyuan and Liu, Tiyao and Zhao, Yawu and Wang, Yuchen and Cheng, Xiaochun},
  title = {PM-DUnet: Fusing long-range dependencies and attention in a dual-U architecture for thyroid nodule segmentation},
  journal = {PLOS One},
  year = {2026},
  volume = {21},
  number = {7},
  pages = {e0353684},
  doi = {10.1371/journal.pone.0353684}
}
Kostenlos starten →