EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

ESD-VesNet: uncertainty-aware vessel segmentation network for endoscopic submucosal dissection with hard negative mining

Mengya Xu, Ming Chen, Zhen Li, Chaoyang Lyu, An Wang, Rulin Zhou, Chuanhao Zhao, Jiaxun Xiang, Tsz Chun Wong, Hossein Farahnaki, Sobhan Zamani Kiasari, Tong Wu und 8 weitere

International Journal of Computer Assisted Radiology and Surgery · 03.07.2026 · 21 (11) · 2343-2354

DOI: 10.1007/s11548-026-03703-2 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Computer Assisted Radiology and Surgery, online erschienen am 03.07.2026. Verfasst von Mengya Xu und 19 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science and Engineering, CUHK, Hong Kong, China. mengyaxu@cuhk.edu.hk. Hauptthemen laut MEDLINE: Endoscopic Mucosal Resection, Blood Loss, Surgical. Die Arbeit nennt 19 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Hong Kong Research Grants Council?(RGC) collaborative Research Fund, General Research Fund, nzhen-Hong Kong-Macau Technology Research Programme (Type C).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Purpose

Preventing intraoperative bleeding in endoscopic submucosal dissection (ESD) relies on the timely identification and pre-coagulation of submucosal vessels. We aim to reduce bleeding risk by delivering accurate vessel segmentation with uncertainty awareness.

Methods

We present ESD-VesNet, a prompt-free SAM3-based framework with an evidential head that outputs vessel probability and pixel-wise uncertainty. The model is trained in a false-positive-aware manner using our developed ESD-Vessel dataset, which includes 2401 positives (images with clear vessel annotation) and 708 curated hard negatives (images may be confused with vessels), combined with online uncertainty-guided hard negative mining. The model is trained in a false-positive-aware manner using the ESD-Vessel dataset, which includes 2401 positive samples (images with clear vessel annotations) and 708 curated hard negatives (images visually confusable with vessels), combined with online uncertainty-guided hard negative mining. By suppressing false positive predictions, our approach has the potential to minimize unnecessary interventions, such as unwarranted coagulation.

Results

On the validation set, ESD-VesNet achieves Dice $$_{FG}$$ FG 52.11%, IoU $$_{FG}$$ FG 35.23%, mDice 75.70%, and Vessel Detection Rate 96.81%, while keeping Background False Positive Rate at 0.61%. Structural quality is high (E-measure 0.7275; S-measure 0.7223). Qualitative analyses show robustness under bubbles, debris, reflections, and blood seepage.

Conclusion

Integrating SAM3 with evidential uncertainty and hard negative mining yields a clinically oriented vessel segmentation system that maintains high detection sensitivity with low false positives, supporting safer ESD.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Endoscopic Mucosal ResectionBlood Loss, SurgicalHumansUncertaintyGastric MucosaData Mining

Stichwörter der Autor:innen

Endoscopic submucosal dissectionEvidential deep learningSAM3Uncertainty estimationVessel segmentation

Autor:innen und Institutionen

  • Department of Computer Science and Engineering, CUHK, Hong Kong, China. mengyaxu@cuhk.edu.hk.
  • Shenzhen Research Institute, CUHK, Shenzhen, China. mengyaxu@cuhk.edu.hk.
  • Department of Electronic Engineering, CUHK, Hong Kong, China.
  • Department of Gastroenterology, Qilu Hospital of Shandong University, Jinan, China.
  • Department of Computer Science and Engineering, CUHK, Hong Kong, China.
  • Department of Electronic Engineering, CUHK, Hong Kong, China. hlren@ieee.org.

Förderung

  • Hong Kong Research Grants Council?(RGC) collaborative Research Fund
  • General Research Fund
  • nzhen-Hong Kong-Macau Technology Research Programme (Type C)

Zitieren

APA

Xu, M., Chen, M., Li, Z., Lyu, C., Wang, A., Zhou, R., Zhao, C., Xiang, J., Wong, T. C., Farahnaki, H., Kiasari, S. Z., Wu, T., Su, Z., Zeng, Y., Wen, R., Shang, X., Mu, Y., Lin, K., Zhang, Y., & Ren, H. (2026). ESD-VesNet: uncertainty-aware vessel segmentation network for endoscopic submucosal dissection with hard negative mining. International Journal of Computer Assisted Radiology and Surgery, 21(11), 2343-2354. https://doi.org/10.1007/s11548-026-03703-2

Vancouver

Xu M, Chen M, Li Z, Lyu C, Wang A, Zhou R, et al. ESD-VesNet: uncertainty-aware vessel segmentation network for endoscopic submucosal dissection with hard negative mining. International Journal of Computer Assisted Radiology and Surgery. 2026;21(11):2343-2354. doi:10.1007/s11548-026-03703-2

BibTeX
@article{Xu2026,
  author = {Xu, Mengya and Chen, Ming and Li, Zhen and Lyu, Chaoyang and Wang, An and Zhou, Rulin and Zhao, Chuanhao and Xiang, Jiaxun and Wong, Tsz Chun and Farahnaki, Hossein and Kiasari, Sobhan Zamani and Wu, Tong and Su, Zimeng and Zeng, Yile and Wen, Ruijing and Shang, Xiaohan and Mu, Yi and Lin, Kezhen and Zhang, Yidong and Ren, Hongliang},
  title = {ESD-VesNet: uncertainty-aware vessel segmentation network for endoscopic submucosal dissection with hard negative mining},
  journal = {International Journal of Computer Assisted Radiology and Surgery},
  year = {2026},
  volume = {21},
  number = {11},
  pages = {2343--2354},
  doi = {10.1007/s11548-026-03703-2}
}
Kostenlos starten →