EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Adaptive sensitivity-fisher regularization for heterogeneous transfer learning of vascular segmentation in laparoscopic videos

Xinkai Zhao, Yuichiro Hayashi, Masahiro Oda, Takayuki Kitasaka, Kazunari Misawa, Kensaku Mori

International Journal of Computer Assisted Radiology and Surgery · 2025

DOI: 10.1007/s11548-025-03404-2 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Computer Assisted Radiology and Surgery, 2025. Verfasst von Xinkai Zhao und 5 weiteren Autor:innen. Beteiligte Einrichtung: Graduate School of Informatics, Nagoya University, Furo-cho, Chikusaku, Nagoya, Aichi, Japan. xkzhao@mori.m.is.nagoya-u.ac.jp. Hauptthemen laut MEDLINE: Laparoscopy, Video Recording, Neural Networks, Computer, Machine Learning. Die Arbeit nennt 32 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Japan Society for the Promotion of Science, Japan Science and Technology Agency, Nagoya University.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Purpose

This study aims to enhance surgical safety by developing a method for vascular segmentation in laparoscopic surgery videos with limited visibility. We introduce an adaptive sensitivity-fisher regularization (ASFR) approach to adapt neural networks, initially trained on non-medical datasets, for vascular segmentation in laparoscopic videos.

Methods

Our approach utilizes heterogeneous transfer learning by integrating fisher information and sensitivity analysis to mitigate catastrophic forgetting and overfitting caused by limited annotated data in laparoscopic videos. We calculate fisher information to identify and preserve critical model parameters while using sensitivity measures to guide adjustment for new task.

Results

The fine-tuned models demonstrated high accuracy in vascular segmentation across various complex video sequences, including those with obscured vessels. For both invisible and visible vessels, our method achieved an average Dice score of 41.3. In addition to outperforming traditional transfer learning approaches, our method exhibited strong adaptability across multiple advanced video segmentation architectures.

Conclusion

This study introduces a novel heterogeneous transfer learning approach, ASFR, which significantly enhances the precision of vascular segmentation in laparoscopic videos. ASFR effectively addresses critical challenges in surgical image analysis and paves the way for broader applications in laparoscopic surgery, promising improved patient outcomes and increased surgical efficiency.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

LaparoscopyVideo RecordingNeural Networks, ComputerMachine LearningImage Processing, Computer-AssistedHumans

Stichwörter der Autor:innen

Laparoscopic surgeryTransfer learningVascular segmentation

Autor:innen und Institutionen

  • Graduate School of Informatics, Nagoya University, Furo-cho, Chikusaku, Nagoya, Aichi, Japan. xkzhao@mori.m.is.nagoya-u.ac.jp.
  • Graduate School of Informatics, Nagoya University, Furo-cho, Chikusaku, Nagoya, Aichi, Japan.
  • Information Technology Center, Nagoya University, Furo-cho, Chikusaku, Nagoya, Aichi, Japan.
  • School of Information Science, Aichi Institute of Technology, 1247 Yachigusa, Yagasa-cho, Toyota, Aichi, Japan.
  • Aichi Cancer Center, 1-1 Kanokoden, Chikusa-ku, Nagoya, Aichi, Japan.
  • Graduate School of Informatics, Nagoya University, Furo-cho, Chikusaku, Nagoya, Aichi, Japan. kensaku@is.nagoya-u.ac.jp.

Förderung

  • Japan Society for the Promotion of Science
  • Japan Science and Technology Agency
  • Nagoya University

Zitieren

APA

Zhao, X., Hayashi, Y., Oda, M., Kitasaka, T., Misawa, K., & Mori, K. (2025). Adaptive sensitivity-fisher regularization for heterogeneous transfer learning of vascular segmentation in laparoscopic videos. International Journal of Computer Assisted Radiology and Surgery. https://doi.org/10.1007/s11548-025-03404-2

Vancouver

Zhao X, Hayashi Y, Oda M, Kitasaka T, Misawa K, Mori K. Adaptive sensitivity-fisher regularization for heterogeneous transfer learning of vascular segmentation in laparoscopic videos. International Journal of Computer Assisted Radiology and Surgery. 2025. doi:10.1007/s11548-025-03404-2

BibTeX
@article{Zhao2025,
  author = {Zhao, Xinkai and Hayashi, Yuichiro and Oda, Masahiro and Kitasaka, Takayuki and Misawa, Kazunari and Mori, Kensaku},
  title = {Adaptive sensitivity-fisher regularization for heterogeneous transfer learning of vascular segmentation in laparoscopic videos},
  journal = {International Journal of Computer Assisted Radiology and Surgery},
  year = {2025},
  doi = {10.1007/s11548-025-03404-2}
}
Kostenlos starten →