EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Stage-Aware Swin-Enhanced nnU-Net v2 for Robust Polyp Segmentation in Collaborative Endoscopic Visual Sensing

Yang Bai, Huan Liu iD

Sensors · 17.08.2026 · 26 (16) · 5206

DOI: 10.3390/s26165206 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 17.08.2026. Verfasst von Yang Bai und 1 weiteren Autor:innen. Beteiligte Einrichtung: School of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150080, China. Hauptthemen laut MEDLINE: Colonic Polyps, Image Processing, Computer-Assisted, Endoscopy. Die Arbeit nennt 30 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Education Department of Heilongjiang Province, Harbin University of Science and Technology, Research Project on Higher Education Teaching Reform in Heilongjiang Province, China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Automatic colon polyp segmentation is important for reliable endoscopic visual sensing. However, segmentation performance can be degraded by ambiguous lesion boundaries, heterogeneous appearance, and image quality variations during acquisition or transmission. This study proposes a stage-aware Swin-enhanced nnU-Net v2 framework, where Swin Transformer blocks are inserted into selected encoder stages while preserving the original nnU-Net v2 pipeline. Different insertion strategies were evaluated on an independent Kvasir-SEG test set, and robustness was further assessed under six synthetic corruption types with three severity levels. The results show that the insertion stage strongly influences the effectiveness of Swin enhancement. Among the evaluated variants, Stage5-Swin achieved the best overall trade-off between segmentation performance, robustness, and computational cost. It achieved the highest Dice scores across all corruption–severity combinations while maintaining comparable external-domain performance on CVC-ClinicDB. Additional FedAvg experiments demonstrated the compatibility of the proposed architecture with collaborative training workflows. Resource analysis further quantified the computational and communication overhead. The findings indicate that middle-to-deep encoder insertion provides a favorable balance between contextual modeling, spatial representation, and efficiency for robust endoscopic segmentation. However, improvements were metric- and corruption-dependent, and moderate overexposure revealed a Precision–Recall trade-off; practical deployment also remains to be validated.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Colonic PolypsImage Processing, Computer-AssistedEndoscopyHumansAlgorithmsNeural Networks, Computer

Stichwörter der Autor:innen

Swin Transformercolon polyp segmentationendoscopic sensingfederated learningimage degradationnnU-Net v2robustnessstage-aware insertion

Autor:innen und Institutionen

  • School of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150080, China.

ORCID-Profile: Huan Liu

Förderung

  • Education Department of Heilongjiang Province
  • Harbin University of Science and Technology
  • Research Project on Higher Education Teaching Reform in Heilongjiang Province, China
  • 2025 Harbin University of Science and Technology Innovation and Entrepreneurship Training Program for Undergraduatesy

Zitieren

APA

Bai, Y., & Liu, H. (2026). Stage-Aware Swin-Enhanced nnU-Net v2 for Robust Polyp Segmentation in Collaborative Endoscopic Visual Sensing. Sensors, 26(16), 5206. https://doi.org/10.3390/s26165206

Vancouver

Bai Y, Liu H. Stage-Aware Swin-Enhanced nnU-Net v2 for Robust Polyp Segmentation in Collaborative Endoscopic Visual Sensing. Sensors. 2026;26(16):5206. doi:10.3390/s26165206

BibTeX
@article{Bai2026,
  author = {Bai, Yang and Liu, Huan},
  title = {Stage-Aware Swin-Enhanced nnU-Net v2 for Robust Polyp Segmentation in Collaborative Endoscopic Visual Sensing},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {16},
  pages = {5206},
  doi = {10.3390/s26165206}
}
Kostenlos starten →