EUVIMEDEUVIMED entdecken
Crossref

Rethinking causality-driven robot tool segmentation with temporal constraints

Hao Ding, Jie Ying Wu, Zhaoshuo Li, Mathias Unberath

International Journal of Computer Assisted Radiology and Surgery · 07.04.2023 · 18 (6) · 1009-1016

DOI: 10.1007/s11548-023-02872-8 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Computer Assisted Radiology and Surgery, online erschienen am 07.04.2023. Verfasst von Hao Ding und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science, Johns Hopkins University, 3400 N. Charles St, Baltimore, MD, 21218, USA. hding15@jhu.edu. Hauptthemen laut MEDLINE: Neural Networks, Computer, Robotics. Die Arbeit wurde bisher 8-mal zitiert (Crossref) und nennt 31 Literaturangaben. Gefördert von: Chinese University of Hong Kong.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

PURPOSE: Vision-based robot tool segmentation plays a fundamental role in surgical robots perception and downstream tasks. CaRTS, based on a complementary causal model, has shown promising performance in unseen counterfactual surgical environments in the presence of smoke, blood, etc. However, CaRTS requires over 30 iterations of optimization to converge for a single image due to limited observability. METHOD: To address the above limitations, we take temporal relation into consideration and propose a temporal causal model for robot tool segmentation on video sequences. We design an architecture named Temporally Constrained CaRTS (TC-CaRTS). TC-CaRTS has three novel modules to complement CaRTS-temporal optimization pipeline, kinematics correction network, and spatial-temporal regularization. RESULTS: Experiment results show that TC-CaRTS requires fewer iterations to achieve the same or better performance as CaRTS on different domains. All three modules are proven to be effective. CONCLUSION: We propose TC-CaRTS, which takes advantage of temporal constraints as additional observability. We show that TC-CaRTS outperforms prior work in the robot tool segmentation task with improved convergence speed on test datasets from different domains.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Neural Networks, ComputerRoboticsHumansBiomechanical PhenomenaImage Processing, Computer-Assisted

Stichwörter der Autor:innen

Computer visionComputer-assisted surgeryDeep learningMinimally invasive surgeryRobustness

Autor:innen und Institutionen

  • Department of Computer Science, Johns Hopkins University, 3400 N. Charles St, Baltimore, MD, 21218, USA. hding15@jhu.edu.
  • Department of Computer Science, Vanderbilt University, 2201 West End Ave, Nashville, TN, 37235, USA.
  • Department of Computer Science, Johns Hopkins University, 3400 N. Charles St, Baltimore, MD, 21218, USA.
  • Department of Computer Science, Johns Hopkins University, 3400 N. Charles St, Baltimore, MD, 21218, USA. unberath@jhu.edu.

Förderung

  • Chinese University of Hong Kong

Zitieren

APA

Ding, H., Wu, J. Y., Li, Z., & Unberath, M. (2023). Rethinking causality-driven robot tool segmentation with temporal constraints. International Journal of Computer Assisted Radiology and Surgery, 18(6), 1009-1016. https://doi.org/10.1007/s11548-023-02872-8

Vancouver

Ding H, Wu JY, Li Z, Unberath M. Rethinking causality-driven robot tool segmentation with temporal constraints. International Journal of Computer Assisted Radiology and Surgery. 2023;18(6):1009-1016. doi:10.1007/s11548-023-02872-8

BibTeX
@article{Ding2023,
  author = {Ding, Hao and Wu, Jie Ying and Li, Zhaoshuo and Unberath, Mathias},
  title = {Rethinking causality-driven robot tool segmentation with temporal constraints},
  journal = {International Journal of Computer Assisted Radiology and Surgery},
  year = {2023},
  volume = {18},
  number = {6},
  pages = {1009--1016},
  doi = {10.1007/s11548-023-02872-8}
}
Kostenlos starten →