Just‐In‐Time Modelling of Real‐Time Tissue Deformation in Ultrasound‐Guided Percutaneous Surgery
International Journal for Numerical Methods in Biomedical Engineering · 2026-11 · 42 (11)
DOI: 10.1002/cnm.70235 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal for Numerical Methods in Biomedical Engineering, online erschienen am 17.11.2026. Verfasst von Kian Zalzalah und 1 weiteren Autor:innen. Beteiligte Einrichtung: Department of Systems and Computer Engineering, Carleton University, Ottawa, Ontario, Canada. Hauptthemen laut MEDLINE: Surgery, Computer-Assisted. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Natural Sciences and Engineering Research Council of Canada.
Abstract (Originalsprache)
Simulators for training ultrasound‐guided percutaneous surgery (usPS) demand rapid updates of tissue and image deformation, yet high‐fidelity physics models exceed real‐time budgets on standard hardware. This paper presents a Just‐in‐Time (JIT) modelling framework that predicts ultrasound image deformation from needle insertions in real‐time. First, an offline historical database of needle locations and resulting image deformations is created using a simulation engine. For each new query, the model selects a local neighbourhood and computes weights that minimize a mean‐square error. The resulting weighted prediction of the full nodal deformation is reshaped for direct simulation or visualization. The framework is evaluated against mesh deformations calculated using thin plate spline (TPS), five mesh resolutions and multiple database sizes. On the finest mesh, the mean relative error is 0.2% while the median per‐query runtime is 1.1 ms, providing a speedup of 31,900 times over the simulation engine used to generate the offline database. These results indicate that JIT closely reproduces the TPS reference deformation while satisfying stringent real‐time requirements. The proposed framework is suitable for real‐time, embedded and portable usPS training simulators.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Systems and Computer Engineering, Carleton University, Ottawa, Ontario, Canada.
ORCID-Profile: Kian Zalzalah
Förderung
- Natural Sciences and Engineering Research Council of Canada
Zitieren
APA
Zalzalah, K., & Rossa, C. (2026). Just‐In‐Time Modelling of Real‐Time Tissue Deformation in Ultrasound‐Guided Percutaneous Surgery. International Journal for Numerical Methods in Biomedical Engineering, 42(11). https://doi.org/10.1002/cnm.70235
Vancouver
Zalzalah K, Rossa C. Just‐In‐Time Modelling of Real‐Time Tissue Deformation in Ultrasound‐Guided Percutaneous Surgery. International Journal for Numerical Methods in Biomedical Engineering. 2026;42(11). doi:10.1002/cnm.70235
BibTeX
@article{Zalzalah2026,
author = {Zalzalah, Kian and Rossa, Carlos},
title = {Just‐In‐Time Modelling of Real‐Time Tissue Deformation in Ultrasound‐Guided Percutaneous Surgery},
journal = {International Journal for Numerical Methods in Biomedical Engineering},
year = {2026},
volume = {42},
number = {11},
doi = {10.1002/cnm.70235}
}
EUVIMED